diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,84021 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.5287597936174278,
+  "eval_steps": 500,
+  "global_step": 12000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.736969857569419,
+      "learning_rate": 2.1231422505307857e-08,
+      "loss": 0.6756,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.037470655132517,
+      "learning_rate": 4.246284501061571e-08,
+      "loss": 0.8562,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9218059224199386,
+      "learning_rate": 6.369426751592358e-08,
+      "loss": 0.8068,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.619998245287453,
+      "learning_rate": 8.492569002123143e-08,
+      "loss": 0.6641,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7552286831643211,
+      "learning_rate": 1.0615711252653928e-07,
+      "loss": 0.6532,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.725938522012392,
+      "learning_rate": 1.2738853503184715e-07,
+      "loss": 0.6431,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8840277525393838,
+      "learning_rate": 1.4861995753715502e-07,
+      "loss": 0.7829,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.787761961910525,
+      "learning_rate": 1.6985138004246285e-07,
+      "loss": 0.6583,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.856382702441212,
+      "learning_rate": 1.9108280254777072e-07,
+      "loss": 0.8299,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9440185036397162,
+      "learning_rate": 2.1231422505307855e-07,
+      "loss": 0.7814,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.042243550306094,
+      "learning_rate": 2.3354564755838642e-07,
+      "loss": 0.8368,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7996395106043221,
+      "learning_rate": 2.547770700636943e-07,
+      "loss": 0.638,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.869579841578725,
+      "learning_rate": 2.7600849256900214e-07,
+      "loss": 0.8517,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9095070756184216,
+      "learning_rate": 2.9723991507431003e-07,
+      "loss": 0.8669,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7011520507056614,
+      "learning_rate": 3.1847133757961787e-07,
+      "loss": 0.6481,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.72656352612305,
+      "learning_rate": 3.397027600849257e-07,
+      "loss": 0.6676,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9850011594794108,
+      "learning_rate": 3.6093418259023354e-07,
+      "loss": 0.8064,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9961052081277577,
+      "learning_rate": 3.8216560509554143e-07,
+      "loss": 0.8623,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5419972208334418,
+      "learning_rate": 4.0339702760084927e-07,
+      "loss": 0.6767,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1838794830736905,
+      "learning_rate": 4.246284501061571e-07,
+      "loss": 0.8244,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9317232935018747,
+      "learning_rate": 4.45859872611465e-07,
+      "loss": 0.852,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7085361334482871,
+      "learning_rate": 4.6709129511677283e-07,
+      "loss": 0.657,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8870619473509618,
+      "learning_rate": 4.883227176220808e-07,
+      "loss": 0.8576,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5299893741179431,
+      "learning_rate": 5.095541401273886e-07,
+      "loss": 0.6873,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4250924984224744,
+      "learning_rate": 5.307855626326964e-07,
+      "loss": 0.7369,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7077717936168544,
+      "learning_rate": 5.520169851380043e-07,
+      "loss": 0.815,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5997246770566833,
+      "learning_rate": 5.732484076433121e-07,
+      "loss": 0.7909,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4952396677004043,
+      "learning_rate": 5.944798301486201e-07,
+      "loss": 0.6704,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3899327550048044,
+      "learning_rate": 6.157112526539279e-07,
+      "loss": 0.6509,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.061712463091723,
+      "learning_rate": 6.369426751592357e-07,
+      "loss": 0.85,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2744370728062344,
+      "learning_rate": 6.581740976645436e-07,
+      "loss": 0.6423,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3923514258885368,
+      "learning_rate": 6.794055201698514e-07,
+      "loss": 0.6969,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3692117861966746,
+      "learning_rate": 7.006369426751592e-07,
+      "loss": 0.6553,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.096917613406094,
+      "learning_rate": 7.218683651804671e-07,
+      "loss": 0.5882,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.23589373051139,
+      "learning_rate": 7.43099787685775e-07,
+      "loss": 0.6099,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.457426821981949,
+      "learning_rate": 7.643312101910829e-07,
+      "loss": 0.8656,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3463087038344559,
+      "learning_rate": 7.855626326963907e-07,
+      "loss": 0.7443,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2354215190204334,
+      "learning_rate": 8.067940552016985e-07,
+      "loss": 0.6589,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3543046797702938,
+      "learning_rate": 8.280254777070064e-07,
+      "loss": 0.7884,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9978970211744808,
+      "learning_rate": 8.492569002123142e-07,
+      "loss": 0.646,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.198186746771855,
+      "learning_rate": 8.70488322717622e-07,
+      "loss": 0.8023,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9222935238354153,
+      "learning_rate": 8.9171974522293e-07,
+      "loss": 0.7114,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7987781325353095,
+      "learning_rate": 9.129511677282378e-07,
+      "loss": 0.5467,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9526224105233752,
+      "learning_rate": 9.341825902335457e-07,
+      "loss": 0.5956,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4331820306922685,
+      "learning_rate": 9.554140127388537e-07,
+      "loss": 0.8467,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8270249691131161,
+      "learning_rate": 9.766454352441615e-07,
+      "loss": 0.6266,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8966224676323107,
+      "learning_rate": 9.978768577494694e-07,
+      "loss": 0.6487,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1974556124489195,
+      "learning_rate": 1.0191082802547772e-06,
+      "loss": 0.8419,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2300556285046356,
+      "learning_rate": 1.040339702760085e-06,
+      "loss": 0.7656,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.088997811624343,
+      "learning_rate": 1.0615711252653929e-06,
+      "loss": 0.6085,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1661197607504696,
+      "learning_rate": 1.0828025477707007e-06,
+      "loss": 0.7957,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8565785576485809,
+      "learning_rate": 1.1040339702760086e-06,
+      "loss": 0.5701,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8073850980562628,
+      "learning_rate": 1.1252653927813164e-06,
+      "loss": 0.6135,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1524349803707088,
+      "learning_rate": 1.1464968152866242e-06,
+      "loss": 0.7108,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8949503229969369,
+      "learning_rate": 1.167728237791932e-06,
+      "loss": 0.644,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1290110901645551,
+      "learning_rate": 1.1889596602972401e-06,
+      "loss": 0.7598,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2260917734247248,
+      "learning_rate": 1.210191082802548e-06,
+      "loss": 0.7976,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7484536915279122,
+      "learning_rate": 1.2314225053078558e-06,
+      "loss": 0.6533,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7018104782389608,
+      "learning_rate": 1.2526539278131636e-06,
+      "loss": 0.6257,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7496189310231463,
+      "learning_rate": 1.2738853503184715e-06,
+      "loss": 0.6163,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0275015023681622,
+      "learning_rate": 1.2951167728237793e-06,
+      "loss": 0.7835,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8049740996421182,
+      "learning_rate": 1.3163481953290871e-06,
+      "loss": 0.5951,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.769250999670833,
+      "learning_rate": 1.337579617834395e-06,
+      "loss": 0.6993,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7367855450371175,
+      "learning_rate": 1.3588110403397028e-06,
+      "loss": 0.5676,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0538913639692464,
+      "learning_rate": 1.3800424628450107e-06,
+      "loss": 0.7137,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7836353836094159,
+      "learning_rate": 1.4012738853503185e-06,
+      "loss": 0.6458,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2681341915414426,
+      "learning_rate": 1.4225053078556263e-06,
+      "loss": 0.7745,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7464717867766247,
+      "learning_rate": 1.4437367303609342e-06,
+      "loss": 0.6341,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7034007234369021,
+      "learning_rate": 1.4649681528662422e-06,
+      "loss": 0.5941,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9151609867840141,
+      "learning_rate": 1.48619957537155e-06,
+      "loss": 0.7036,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8111596902625513,
+      "learning_rate": 1.5074309978768579e-06,
+      "loss": 0.6062,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7583995233620089,
+      "learning_rate": 1.5286624203821657e-06,
+      "loss": 0.5653,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9962891903702115,
+      "learning_rate": 1.5498938428874736e-06,
+      "loss": 0.7762,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1313304547472938,
+      "learning_rate": 1.5711252653927814e-06,
+      "loss": 0.7421,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.008144720140337,
+      "learning_rate": 1.5923566878980892e-06,
+      "loss": 0.7189,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7780427717780393,
+      "learning_rate": 1.613588110403397e-06,
+      "loss": 0.5921,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8399311727829962,
+      "learning_rate": 1.634819532908705e-06,
+      "loss": 0.5997,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6927282906786248,
+      "learning_rate": 1.6560509554140127e-06,
+      "loss": 0.5671,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2300098877330876,
+      "learning_rate": 1.6772823779193206e-06,
+      "loss": 0.7011,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9185988906241528,
+      "learning_rate": 1.6985138004246284e-06,
+      "loss": 0.6905,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9206959465561452,
+      "learning_rate": 1.7197452229299363e-06,
+      "loss": 0.7107,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6818779957755264,
+      "learning_rate": 1.740976645435244e-06,
+      "loss": 0.5602,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6498799279522858,
+      "learning_rate": 1.7622080679405521e-06,
+      "loss": 0.601,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.888398773343403,
+      "learning_rate": 1.78343949044586e-06,
+      "loss": 0.6146,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.962512792773225,
+      "learning_rate": 1.8046709129511678e-06,
+      "loss": 0.726,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.14206816976848,
+      "learning_rate": 1.8259023354564756e-06,
+      "loss": 0.7984,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8072834925084155,
+      "learning_rate": 1.8471337579617835e-06,
+      "loss": 0.7032,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.947602536205962,
+      "learning_rate": 1.8683651804670913e-06,
+      "loss": 0.6548,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9143970421375164,
+      "learning_rate": 1.8895966029723994e-06,
+      "loss": 0.7302,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.674175642032997,
+      "learning_rate": 1.9108280254777074e-06,
+      "loss": 0.5131,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7886837529805113,
+      "learning_rate": 1.9320594479830153e-06,
+      "loss": 0.7306,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8998842154200234,
+      "learning_rate": 1.953290870488323e-06,
+      "loss": 0.7619,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8939754429002351,
+      "learning_rate": 1.974522292993631e-06,
+      "loss": 0.6977,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7267793453096345,
+      "learning_rate": 1.9957537154989388e-06,
+      "loss": 0.6719,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9423699329896398,
+      "learning_rate": 2.0169851380042466e-06,
+      "loss": 0.6696,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9127968272713965,
+      "learning_rate": 2.0382165605095544e-06,
+      "loss": 0.6863,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6453158741901237,
+      "learning_rate": 2.0594479830148623e-06,
+      "loss": 0.5726,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8004608999355025,
+      "learning_rate": 2.08067940552017e-06,
+      "loss": 0.6646,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.741422175795783,
+      "learning_rate": 2.101910828025478e-06,
+      "loss": 0.6141,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6185567152561202,
+      "learning_rate": 2.1231422505307858e-06,
+      "loss": 0.6171,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.912129457483921,
+      "learning_rate": 2.1443736730360936e-06,
+      "loss": 0.7509,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8247609672782598,
+      "learning_rate": 2.1656050955414015e-06,
+      "loss": 0.6468,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8029055653595574,
+      "learning_rate": 2.1868365180467093e-06,
+      "loss": 0.7296,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9851633577727037,
+      "learning_rate": 2.208067940552017e-06,
+      "loss": 0.744,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6679860577610721,
+      "learning_rate": 2.229299363057325e-06,
+      "loss": 0.5683,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.819612351171737,
+      "learning_rate": 2.250530785562633e-06,
+      "loss": 0.6859,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8845853691655635,
+      "learning_rate": 2.2717622080679406e-06,
+      "loss": 0.7128,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9896458089509722,
+      "learning_rate": 2.2929936305732485e-06,
+      "loss": 0.6823,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8401401133737159,
+      "learning_rate": 2.3142250530785563e-06,
+      "loss": 0.6848,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7044845791337637,
+      "learning_rate": 2.335456475583864e-06,
+      "loss": 0.6287,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8936501983713924,
+      "learning_rate": 2.356687898089172e-06,
+      "loss": 0.6301,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7977858105718107,
+      "learning_rate": 2.3779193205944802e-06,
+      "loss": 0.6501,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9625024469697634,
+      "learning_rate": 2.399150743099788e-06,
+      "loss": 0.7136,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8202496777053818,
+      "learning_rate": 2.420382165605096e-06,
+      "loss": 0.6466,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8083756378016473,
+      "learning_rate": 2.4416135881104038e-06,
+      "loss": 0.5758,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6794575896854108,
+      "learning_rate": 2.4628450106157116e-06,
+      "loss": 0.6271,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.920475184785038,
+      "learning_rate": 2.4840764331210194e-06,
+      "loss": 0.6896,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0624376066131578,
+      "learning_rate": 2.5053078556263273e-06,
+      "loss": 0.6885,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3494780155295418,
+      "learning_rate": 2.526539278131635e-06,
+      "loss": 0.6805,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6680598812484428,
+      "learning_rate": 2.547770700636943e-06,
+      "loss": 0.6003,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7145951603724503,
+      "learning_rate": 2.5690021231422508e-06,
+      "loss": 0.5845,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.932029401239929,
+      "learning_rate": 2.5902335456475586e-06,
+      "loss": 0.7106,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6711551268440451,
+      "learning_rate": 2.6114649681528665e-06,
+      "loss": 0.5839,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8123840964518879,
+      "learning_rate": 2.6326963906581743e-06,
+      "loss": 0.6535,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9502589598494059,
+      "learning_rate": 2.653927813163482e-06,
+      "loss": 0.6723,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.686260030484047,
+      "learning_rate": 2.67515923566879e-06,
+      "loss": 0.5849,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.671346829133467,
+      "learning_rate": 2.696390658174098e-06,
+      "loss": 0.5539,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7348439508382469,
+      "learning_rate": 2.7176220806794056e-06,
+      "loss": 0.6279,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6657544130751492,
+      "learning_rate": 2.7388535031847135e-06,
+      "loss": 0.5721,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7505039321902811,
+      "learning_rate": 2.7600849256900213e-06,
+      "loss": 0.6523,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9081812514026412,
+      "learning_rate": 2.781316348195329e-06,
+      "loss": 0.7182,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1754570547070735,
+      "learning_rate": 2.802547770700637e-06,
+      "loss": 0.7845,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.759204923968475,
+      "learning_rate": 2.823779193205945e-06,
+      "loss": 0.5666,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9134266282272141,
+      "learning_rate": 2.8450106157112527e-06,
+      "loss": 0.617,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6743132459299028,
+      "learning_rate": 2.8662420382165605e-06,
+      "loss": 0.5926,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2852422502660135,
+      "learning_rate": 2.8874734607218683e-06,
+      "loss": 0.7055,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0032638917986518,
+      "learning_rate": 2.908704883227176e-06,
+      "loss": 0.6845,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6545452001482567,
+      "learning_rate": 2.9299363057324844e-06,
+      "loss": 0.5724,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0033165293069863,
+      "learning_rate": 2.9511677282377923e-06,
+      "loss": 0.708,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6603942673614058,
+      "learning_rate": 2.9723991507431e-06,
+      "loss": 0.5816,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9929599681631902,
+      "learning_rate": 2.993630573248408e-06,
+      "loss": 0.652,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9195381630729605,
+      "learning_rate": 3.0148619957537158e-06,
+      "loss": 0.6224,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.96507859825706,
+      "learning_rate": 3.0360934182590236e-06,
+      "loss": 0.7299,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6331253476001903,
+      "learning_rate": 3.0573248407643314e-06,
+      "loss": 0.5364,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7321637764611932,
+      "learning_rate": 3.0785562632696393e-06,
+      "loss": 0.6539,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8029848201541017,
+      "learning_rate": 3.099787685774947e-06,
+      "loss": 0.6919,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8112265854608945,
+      "learning_rate": 3.121019108280255e-06,
+      "loss": 0.5885,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.945927793705502,
+      "learning_rate": 3.142250530785563e-06,
+      "loss": 0.6859,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6415884691227447,
+      "learning_rate": 3.1634819532908706e-06,
+      "loss": 0.5566,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7085088857450645,
+      "learning_rate": 3.1847133757961785e-06,
+      "loss": 0.6582,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7086688356979257,
+      "learning_rate": 3.2059447983014863e-06,
+      "loss": 0.5684,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.89752749128589,
+      "learning_rate": 3.227176220806794e-06,
+      "loss": 0.6836,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7609285062708646,
+      "learning_rate": 3.248407643312102e-06,
+      "loss": 0.6002,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8250009585344907,
+      "learning_rate": 3.26963906581741e-06,
+      "loss": 0.6999,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6295758904840901,
+      "learning_rate": 3.2908704883227177e-06,
+      "loss": 0.6006,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.657018296450642,
+      "learning_rate": 3.3121019108280255e-06,
+      "loss": 0.5709,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7453449119436517,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.6195,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6623503898673194,
+      "learning_rate": 3.354564755838641e-06,
+      "loss": 0.618,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8975443110926178,
+      "learning_rate": 3.375796178343949e-06,
+      "loss": 0.7095,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7196920125346429,
+      "learning_rate": 3.397027600849257e-06,
+      "loss": 0.5944,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7274047246720723,
+      "learning_rate": 3.4182590233545647e-06,
+      "loss": 0.6097,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8416353775608544,
+      "learning_rate": 3.4394904458598725e-06,
+      "loss": 0.6731,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7565544021842142,
+      "learning_rate": 3.4607218683651803e-06,
+      "loss": 0.5338,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8198347197807316,
+      "learning_rate": 3.481953290870488e-06,
+      "loss": 0.5975,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5958765236284512,
+      "learning_rate": 3.5031847133757964e-06,
+      "loss": 0.5593,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8596154311044288,
+      "learning_rate": 3.5244161358811043e-06,
+      "loss": 0.7188,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2250817147186526,
+      "learning_rate": 3.545647558386412e-06,
+      "loss": 0.6885,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6555194561718294,
+      "learning_rate": 3.56687898089172e-06,
+      "loss": 0.584,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.620333165580415,
+      "learning_rate": 3.5881104033970278e-06,
+      "loss": 0.5509,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8073633638541113,
+      "learning_rate": 3.6093418259023356e-06,
+      "loss": 0.5459,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7528890153499365,
+      "learning_rate": 3.6305732484076435e-06,
+      "loss": 0.6095,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9357230716689366,
+      "learning_rate": 3.6518046709129513e-06,
+      "loss": 0.6951,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8441946942232076,
+      "learning_rate": 3.673036093418259e-06,
+      "loss": 0.686,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6953738393442778,
+      "learning_rate": 3.694267515923567e-06,
+      "loss": 0.5587,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0614940309536194,
+      "learning_rate": 3.715498938428875e-06,
+      "loss": 0.6437,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8910635949918921,
+      "learning_rate": 3.7367303609341826e-06,
+      "loss": 0.7033,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8732162877196056,
+      "learning_rate": 3.757961783439491e-06,
+      "loss": 0.6226,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6165029790713727,
+      "learning_rate": 3.7791932059447987e-06,
+      "loss": 0.5568,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7087055049069805,
+      "learning_rate": 3.8004246284501066e-06,
+      "loss": 0.5936,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6614937596481818,
+      "learning_rate": 3.821656050955415e-06,
+      "loss": 0.6021,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7181004064853254,
+      "learning_rate": 3.842887473460722e-06,
+      "loss": 0.5976,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8746688456208674,
+      "learning_rate": 3.8641188959660305e-06,
+      "loss": 0.7057,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8168122830164994,
+      "learning_rate": 3.885350318471338e-06,
+      "loss": 0.6907,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7355109794661671,
+      "learning_rate": 3.906581740976646e-06,
+      "loss": 0.5938,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9557411818018728,
+      "learning_rate": 3.927813163481954e-06,
+      "loss": 0.7442,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8272193627767886,
+      "learning_rate": 3.949044585987262e-06,
+      "loss": 0.7387,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7601867318329757,
+      "learning_rate": 3.970276008492569e-06,
+      "loss": 0.5666,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7848861700574545,
+      "learning_rate": 3.9915074309978775e-06,
+      "loss": 0.6065,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6641571738403048,
+      "learning_rate": 4.012738853503185e-06,
+      "loss": 0.5761,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6167229488978831,
+      "learning_rate": 4.033970276008493e-06,
+      "loss": 0.6061,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9845644834250507,
+      "learning_rate": 4.055201698513801e-06,
+      "loss": 0.6665,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4225166467098442,
+      "learning_rate": 4.076433121019109e-06,
+      "loss": 0.7222,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.825520396357522,
+      "learning_rate": 4.097664543524416e-06,
+      "loss": 0.6237,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9855794534684909,
+      "learning_rate": 4.1188959660297246e-06,
+      "loss": 0.7422,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8740524318615315,
+      "learning_rate": 4.140127388535032e-06,
+      "loss": 0.6686,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7916317776172551,
+      "learning_rate": 4.16135881104034e-06,
+      "loss": 0.6258,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7074471847536283,
+      "learning_rate": 4.1825902335456485e-06,
+      "loss": 0.6024,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7027202324237053,
+      "learning_rate": 4.203821656050956e-06,
+      "loss": 0.5537,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1408714976242633,
+      "learning_rate": 4.225053078556264e-06,
+      "loss": 0.634,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8283598588701209,
+      "learning_rate": 4.2462845010615716e-06,
+      "loss": 0.6226,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9313008468844602,
+      "learning_rate": 4.26751592356688e-06,
+      "loss": 0.7185,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8810462807052284,
+      "learning_rate": 4.288747346072187e-06,
+      "loss": 0.6209,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9015348702685758,
+      "learning_rate": 4.3099787685774955e-06,
+      "loss": 0.6665,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5941279491487401,
+      "learning_rate": 4.331210191082803e-06,
+      "loss": 0.5455,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6242764583731539,
+      "learning_rate": 4.352441613588111e-06,
+      "loss": 0.5779,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7002150286010799,
+      "learning_rate": 4.373673036093419e-06,
+      "loss": 0.5447,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6912798497967817,
+      "learning_rate": 4.394904458598727e-06,
+      "loss": 0.5988,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8690559647733447,
+      "learning_rate": 4.416135881104034e-06,
+      "loss": 0.6324,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0281067404352298,
+      "learning_rate": 4.4373673036093425e-06,
+      "loss": 0.6801,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6895297615817462,
+      "learning_rate": 4.45859872611465e-06,
+      "loss": 0.5542,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6504077645341889,
+      "learning_rate": 4.479830148619958e-06,
+      "loss": 0.5956,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8892541805800651,
+      "learning_rate": 4.501061571125266e-06,
+      "loss": 0.651,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7710720325367937,
+      "learning_rate": 4.522292993630574e-06,
+      "loss": 0.713,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.675110704188669,
+      "learning_rate": 4.543524416135881e-06,
+      "loss": 0.5562,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7072695261962337,
+      "learning_rate": 4.5647558386411895e-06,
+      "loss": 0.6159,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7444640008129021,
+      "learning_rate": 4.585987261146497e-06,
+      "loss": 0.6267,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8671132992439867,
+      "learning_rate": 4.607218683651805e-06,
+      "loss": 0.6657,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7074397549705441,
+      "learning_rate": 4.628450106157113e-06,
+      "loss": 0.558,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7800357800835367,
+      "learning_rate": 4.649681528662421e-06,
+      "loss": 0.5888,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7353721538536665,
+      "learning_rate": 4.670912951167728e-06,
+      "loss": 0.6019,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9013761422335163,
+      "learning_rate": 4.6921443736730366e-06,
+      "loss": 0.636,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7469190290966377,
+      "learning_rate": 4.713375796178344e-06,
+      "loss": 0.5989,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7270094404967548,
+      "learning_rate": 4.734607218683652e-06,
+      "loss": 0.6204,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6216399989244757,
+      "learning_rate": 4.7558386411889605e-06,
+      "loss": 0.5971,
+      "step": 224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6066311313104265,
+      "learning_rate": 4.777070063694268e-06,
+      "loss": 0.6148,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5599471218785727,
+      "learning_rate": 4.798301486199576e-06,
+      "loss": 0.4796,
+      "step": 226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6157753345646437,
+      "learning_rate": 4.819532908704884e-06,
+      "loss": 0.5216,
+      "step": 227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8228195083528653,
+      "learning_rate": 4.840764331210192e-06,
+      "loss": 0.6043,
+      "step": 228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.686772813135059,
+      "learning_rate": 4.861995753715499e-06,
+      "loss": 0.6483,
+      "step": 229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9872220934878595,
+      "learning_rate": 4.8832271762208075e-06,
+      "loss": 0.7057,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9814221220293042,
+      "learning_rate": 4.904458598726115e-06,
+      "loss": 0.6954,
+      "step": 231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.89164011450147,
+      "learning_rate": 4.925690021231423e-06,
+      "loss": 0.5815,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7016484179686268,
+      "learning_rate": 4.946921443736731e-06,
+      "loss": 0.6257,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.774515551923184,
+      "learning_rate": 4.968152866242039e-06,
+      "loss": 0.6715,
+      "step": 234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8137485818123509,
+      "learning_rate": 4.989384288747346e-06,
+      "loss": 0.6076,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7561940959299209,
+      "learning_rate": 5.0106157112526545e-06,
+      "loss": 0.5821,
+      "step": 236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.898623103820379,
+      "learning_rate": 5.031847133757962e-06,
+      "loss": 0.706,
+      "step": 237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7995953466662316,
+      "learning_rate": 5.05307855626327e-06,
+      "loss": 0.6196,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6536283203986453,
+      "learning_rate": 5.074309978768578e-06,
+      "loss": 0.5753,
+      "step": 239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.051355284149499,
+      "learning_rate": 5.095541401273886e-06,
+      "loss": 0.6036,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8991691183020971,
+      "learning_rate": 5.116772823779193e-06,
+      "loss": 0.6782,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6256156762749676,
+      "learning_rate": 5.1380042462845016e-06,
+      "loss": 0.5823,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7177094517309004,
+      "learning_rate": 5.159235668789809e-06,
+      "loss": 0.6223,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8711944375615034,
+      "learning_rate": 5.180467091295117e-06,
+      "loss": 0.6944,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8803187138200502,
+      "learning_rate": 5.201698513800425e-06,
+      "loss": 0.6699,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7775556410870242,
+      "learning_rate": 5.222929936305733e-06,
+      "loss": 0.5807,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8186463739188571,
+      "learning_rate": 5.24416135881104e-06,
+      "loss": 0.6677,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8494691832014075,
+      "learning_rate": 5.265392781316349e-06,
+      "loss": 0.7448,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7690654223032175,
+      "learning_rate": 5.286624203821657e-06,
+      "loss": 0.5999,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6041357286634986,
+      "learning_rate": 5.307855626326964e-06,
+      "loss": 0.55,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6227306599717993,
+      "learning_rate": 5.3290870488322725e-06,
+      "loss": 0.5158,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7241136824678885,
+      "learning_rate": 5.35031847133758e-06,
+      "loss": 0.6279,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6810033718421417,
+      "learning_rate": 5.371549893842888e-06,
+      "loss": 0.5917,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8444944767654462,
+      "learning_rate": 5.392781316348196e-06,
+      "loss": 0.6579,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7122852417532933,
+      "learning_rate": 5.414012738853504e-06,
+      "loss": 0.5785,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6718345646223689,
+      "learning_rate": 5.435244161358811e-06,
+      "loss": 0.5623,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8559619562534222,
+      "learning_rate": 5.4564755838641195e-06,
+      "loss": 0.6424,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9897427075245315,
+      "learning_rate": 5.477707006369427e-06,
+      "loss": 0.6599,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7495990121048346,
+      "learning_rate": 5.498938428874735e-06,
+      "loss": 0.7048,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7617635842160821,
+      "learning_rate": 5.520169851380043e-06,
+      "loss": 0.6453,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9923681855393078,
+      "learning_rate": 5.541401273885351e-06,
+      "loss": 0.7383,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8222305195153744,
+      "learning_rate": 5.562632696390658e-06,
+      "loss": 0.6413,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6626699551761589,
+      "learning_rate": 5.5838641188959666e-06,
+      "loss": 0.5702,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8119419623711125,
+      "learning_rate": 5.605095541401274e-06,
+      "loss": 0.6073,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7647961887972264,
+      "learning_rate": 5.626326963906582e-06,
+      "loss": 0.5806,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6166367579382988,
+      "learning_rate": 5.64755838641189e-06,
+      "loss": 0.5656,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.799875375532442,
+      "learning_rate": 5.668789808917198e-06,
+      "loss": 0.649,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7275526056631297,
+      "learning_rate": 5.690021231422505e-06,
+      "loss": 0.6317,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6472681387536379,
+      "learning_rate": 5.7112526539278136e-06,
+      "loss": 0.5783,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.637531562482795,
+      "learning_rate": 5.732484076433121e-06,
+      "loss": 0.5846,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7397922736008121,
+      "learning_rate": 5.753715498938429e-06,
+      "loss": 0.6684,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0205806869608305,
+      "learning_rate": 5.774946921443737e-06,
+      "loss": 0.6707,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8699718878359272,
+      "learning_rate": 5.796178343949045e-06,
+      "loss": 0.6475,
+      "step": 273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6690727819694017,
+      "learning_rate": 5.817409766454352e-06,
+      "loss": 0.6253,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8550115033966339,
+      "learning_rate": 5.838641188959661e-06,
+      "loss": 0.6021,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8652339155907169,
+      "learning_rate": 5.859872611464969e-06,
+      "loss": 0.6385,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8573757882631059,
+      "learning_rate": 5.881104033970276e-06,
+      "loss": 0.5985,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7481145556491249,
+      "learning_rate": 5.9023354564755845e-06,
+      "loss": 0.5614,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7603850815803599,
+      "learning_rate": 5.923566878980892e-06,
+      "loss": 0.5895,
+      "step": 279
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7106234283638302,
+      "learning_rate": 5.9447983014862e-06,
+      "loss": 0.5841,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7571796686983795,
+      "learning_rate": 5.966029723991508e-06,
+      "loss": 0.5725,
+      "step": 281
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6563732440289131,
+      "learning_rate": 5.987261146496816e-06,
+      "loss": 0.6088,
+      "step": 282
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6977442128338657,
+      "learning_rate": 6.008492569002123e-06,
+      "loss": 0.5594,
+      "step": 283
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.706010792363827,
+      "learning_rate": 6.0297239915074315e-06,
+      "loss": 0.5695,
+      "step": 284
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.859176506136312,
+      "learning_rate": 6.050955414012739e-06,
+      "loss": 0.667,
+      "step": 285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8105147566645884,
+      "learning_rate": 6.072186836518047e-06,
+      "loss": 0.632,
+      "step": 286
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8209217482786072,
+      "learning_rate": 6.093418259023355e-06,
+      "loss": 0.6366,
+      "step": 287
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7608301811315319,
+      "learning_rate": 6.114649681528663e-06,
+      "loss": 0.5933,
+      "step": 288
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7654007541474549,
+      "learning_rate": 6.13588110403397e-06,
+      "loss": 0.6515,
+      "step": 289
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.629612742923365,
+      "learning_rate": 6.1571125265392786e-06,
+      "loss": 0.571,
+      "step": 290
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7279126481255758,
+      "learning_rate": 6.178343949044586e-06,
+      "loss": 0.6631,
+      "step": 291
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6004298020074539,
+      "learning_rate": 6.199575371549894e-06,
+      "loss": 0.5499,
+      "step": 292
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6887789599745987,
+      "learning_rate": 6.220806794055202e-06,
+      "loss": 0.6049,
+      "step": 293
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7972200861030105,
+      "learning_rate": 6.24203821656051e-06,
+      "loss": 0.6428,
+      "step": 294
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7465288393939438,
+      "learning_rate": 6.263269639065817e-06,
+      "loss": 0.587,
+      "step": 295
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0574613089275766,
+      "learning_rate": 6.284501061571126e-06,
+      "loss": 0.6344,
+      "step": 296
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9460877485374782,
+      "learning_rate": 6.305732484076433e-06,
+      "loss": 0.6436,
+      "step": 297
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8184763277187621,
+      "learning_rate": 6.326963906581741e-06,
+      "loss": 0.7271,
+      "step": 298
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6930575103378493,
+      "learning_rate": 6.348195329087049e-06,
+      "loss": 0.6041,
+      "step": 299
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7531353169689462,
+      "learning_rate": 6.369426751592357e-06,
+      "loss": 0.5961,
+      "step": 300
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7100085841084717,
+      "learning_rate": 6.390658174097664e-06,
+      "loss": 0.5595,
+      "step": 301
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9048480632880608,
+      "learning_rate": 6.411889596602973e-06,
+      "loss": 0.6511,
+      "step": 302
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6880597100842921,
+      "learning_rate": 6.433121019108281e-06,
+      "loss": 0.5716,
+      "step": 303
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8482267906196166,
+      "learning_rate": 6.454352441613588e-06,
+      "loss": 0.6337,
+      "step": 304
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7599383356399694,
+      "learning_rate": 6.4755838641188965e-06,
+      "loss": 0.5896,
+      "step": 305
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6976717143745537,
+      "learning_rate": 6.496815286624204e-06,
+      "loss": 0.5877,
+      "step": 306
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9891232937813345,
+      "learning_rate": 6.518046709129512e-06,
+      "loss": 0.741,
+      "step": 307
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7442534546756299,
+      "learning_rate": 6.53927813163482e-06,
+      "loss": 0.6178,
+      "step": 308
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7970249741724695,
+      "learning_rate": 6.560509554140128e-06,
+      "loss": 0.5939,
+      "step": 309
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8354095101493896,
+      "learning_rate": 6.581740976645435e-06,
+      "loss": 0.7177,
+      "step": 310
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6420658221124799,
+      "learning_rate": 6.6029723991507436e-06,
+      "loss": 0.5191,
+      "step": 311
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7228803925846446,
+      "learning_rate": 6.624203821656051e-06,
+      "loss": 0.5701,
+      "step": 312
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7886061379417076,
+      "learning_rate": 6.645435244161359e-06,
+      "loss": 0.6825,
+      "step": 313
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6494021019881504,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.6025,
+      "step": 314
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.064902467884702,
+      "learning_rate": 6.687898089171975e-06,
+      "loss": 0.7276,
+      "step": 315
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7076045071300145,
+      "learning_rate": 6.709129511677282e-06,
+      "loss": 0.5927,
+      "step": 316
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6516034356887127,
+      "learning_rate": 6.730360934182591e-06,
+      "loss": 0.5837,
+      "step": 317
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8532912970246789,
+      "learning_rate": 6.751592356687898e-06,
+      "loss": 0.6974,
+      "step": 318
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8136814955252396,
+      "learning_rate": 6.772823779193206e-06,
+      "loss": 0.6814,
+      "step": 319
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7694784076036806,
+      "learning_rate": 6.794055201698514e-06,
+      "loss": 0.6535,
+      "step": 320
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6080238914599104,
+      "learning_rate": 6.815286624203822e-06,
+      "loss": 0.57,
+      "step": 321
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8143841412568615,
+      "learning_rate": 6.836518046709129e-06,
+      "loss": 0.5867,
+      "step": 322
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9020756728287854,
+      "learning_rate": 6.857749469214438e-06,
+      "loss": 0.6342,
+      "step": 323
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9170186790597973,
+      "learning_rate": 6.878980891719745e-06,
+      "loss": 0.701,
+      "step": 324
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6878753443589771,
+      "learning_rate": 6.900212314225053e-06,
+      "loss": 0.5742,
+      "step": 325
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6721737113608273,
+      "learning_rate": 6.921443736730361e-06,
+      "loss": 0.5557,
+      "step": 326
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6550388119556643,
+      "learning_rate": 6.942675159235669e-06,
+      "loss": 0.6052,
+      "step": 327
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7399479184989896,
+      "learning_rate": 6.963906581740976e-06,
+      "loss": 0.6463,
+      "step": 328
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7156549908824404,
+      "learning_rate": 6.985138004246285e-06,
+      "loss": 0.5728,
+      "step": 329
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6718801324294158,
+      "learning_rate": 7.006369426751593e-06,
+      "loss": 0.5629,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7009074851724983,
+      "learning_rate": 7.0276008492569e-06,
+      "loss": 0.5532,
+      "step": 331
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8029408411971348,
+      "learning_rate": 7.0488322717622086e-06,
+      "loss": 0.5271,
+      "step": 332
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0431012036211642,
+      "learning_rate": 7.070063694267516e-06,
+      "loss": 0.6367,
+      "step": 333
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8093878262347405,
+      "learning_rate": 7.091295116772824e-06,
+      "loss": 0.6403,
+      "step": 334
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2908023804812134,
+      "learning_rate": 7.112526539278132e-06,
+      "loss": 0.7034,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0398537291927146,
+      "learning_rate": 7.13375796178344e-06,
+      "loss": 0.6809,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0350876345922804,
+      "learning_rate": 7.154989384288747e-06,
+      "loss": 0.6931,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7886373173499229,
+      "learning_rate": 7.1762208067940556e-06,
+      "loss": 0.6589,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8927862733820032,
+      "learning_rate": 7.197452229299363e-06,
+      "loss": 0.6935,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8259408573357077,
+      "learning_rate": 7.218683651804671e-06,
+      "loss": 0.6376,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7824414331887,
+      "learning_rate": 7.239915074309979e-06,
+      "loss": 0.5965,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6088715758221711,
+      "learning_rate": 7.261146496815287e-06,
+      "loss": 0.5885,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0709181495605602,
+      "learning_rate": 7.282377919320594e-06,
+      "loss": 0.6809,
+      "step": 343
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8043326912255777,
+      "learning_rate": 7.303609341825903e-06,
+      "loss": 0.6297,
+      "step": 344
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8565291208373391,
+      "learning_rate": 7.32484076433121e-06,
+      "loss": 0.6546,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.624084247670802,
+      "learning_rate": 7.346072186836518e-06,
+      "loss": 0.5563,
+      "step": 346
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7824732109326937,
+      "learning_rate": 7.367303609341826e-06,
+      "loss": 0.6785,
+      "step": 347
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8399482169207614,
+      "learning_rate": 7.388535031847134e-06,
+      "loss": 0.5974,
+      "step": 348
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.785521414538481,
+      "learning_rate": 7.409766454352441e-06,
+      "loss": 0.6213,
+      "step": 349
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8548704922581076,
+      "learning_rate": 7.43099787685775e-06,
+      "loss": 0.6807,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7805367105203607,
+      "learning_rate": 7.452229299363057e-06,
+      "loss": 0.6876,
+      "step": 351
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7911626509097884,
+      "learning_rate": 7.473460721868365e-06,
+      "loss": 0.6201,
+      "step": 352
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9263847215601537,
+      "learning_rate": 7.494692144373673e-06,
+      "loss": 0.6726,
+      "step": 353
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.773292037382418,
+      "learning_rate": 7.515923566878982e-06,
+      "loss": 0.6163,
+      "step": 354
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8488440940786954,
+      "learning_rate": 7.53715498938429e-06,
+      "loss": 0.5657,
+      "step": 355
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8985238040159557,
+      "learning_rate": 7.5583864118895975e-06,
+      "loss": 0.6968,
+      "step": 356
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7376906256571085,
+      "learning_rate": 7.579617834394906e-06,
+      "loss": 0.5766,
+      "step": 357
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6623660432305776,
+      "learning_rate": 7.600849256900213e-06,
+      "loss": 0.5424,
+      "step": 358
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7753384538437651,
+      "learning_rate": 7.622080679405521e-06,
+      "loss": 0.5261,
+      "step": 359
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7636721584433728,
+      "learning_rate": 7.64331210191083e-06,
+      "loss": 0.5604,
+      "step": 360
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.807227120074653,
+      "learning_rate": 7.664543524416136e-06,
+      "loss": 0.6237,
+      "step": 361
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7202077397488372,
+      "learning_rate": 7.685774946921445e-06,
+      "loss": 0.5967,
+      "step": 362
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9238434706684684,
+      "learning_rate": 7.707006369426753e-06,
+      "loss": 0.6284,
+      "step": 363
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6904290880218553,
+      "learning_rate": 7.728237791932061e-06,
+      "loss": 0.5772,
+      "step": 364
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6759959514969673,
+      "learning_rate": 7.74946921443737e-06,
+      "loss": 0.5675,
+      "step": 365
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.575640820985936,
+      "learning_rate": 7.770700636942676e-06,
+      "loss": 0.6411,
+      "step": 366
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.976655319737452,
+      "learning_rate": 7.791932059447984e-06,
+      "loss": 0.7161,
+      "step": 367
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6794250834654887,
+      "learning_rate": 7.813163481953292e-06,
+      "loss": 0.5621,
+      "step": 368
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8958077513340039,
+      "learning_rate": 7.8343949044586e-06,
+      "loss": 0.6572,
+      "step": 369
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7730165050550669,
+      "learning_rate": 7.855626326963907e-06,
+      "loss": 0.6202,
+      "step": 370
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.744079962426877,
+      "learning_rate": 7.876857749469215e-06,
+      "loss": 0.6058,
+      "step": 371
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8540704691717359,
+      "learning_rate": 7.898089171974524e-06,
+      "loss": 0.5719,
+      "step": 372
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8631284956302128,
+      "learning_rate": 7.919320594479832e-06,
+      "loss": 0.6008,
+      "step": 373
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8642598382843337,
+      "learning_rate": 7.940552016985139e-06,
+      "loss": 0.614,
+      "step": 374
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7439967365258471,
+      "learning_rate": 7.961783439490447e-06,
+      "loss": 0.5856,
+      "step": 375
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6795607401111311,
+      "learning_rate": 7.983014861995755e-06,
+      "loss": 0.5841,
+      "step": 376
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7969899454925713,
+      "learning_rate": 8.004246284501063e-06,
+      "loss": 0.6461,
+      "step": 377
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.975299506666238,
+      "learning_rate": 8.02547770700637e-06,
+      "loss": 0.6425,
+      "step": 378
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8174766590854395,
+      "learning_rate": 8.046709129511678e-06,
+      "loss": 0.6229,
+      "step": 379
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6061374519640897,
+      "learning_rate": 8.067940552016986e-06,
+      "loss": 0.5521,
+      "step": 380
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2087893068940179,
+      "learning_rate": 8.089171974522295e-06,
+      "loss": 0.7737,
+      "step": 381
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8486748058122714,
+      "learning_rate": 8.110403397027601e-06,
+      "loss": 0.5821,
+      "step": 382
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6690153487406375,
+      "learning_rate": 8.13163481953291e-06,
+      "loss": 0.5573,
+      "step": 383
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8056415402086693,
+      "learning_rate": 8.152866242038218e-06,
+      "loss": 0.695,
+      "step": 384
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7857803059268238,
+      "learning_rate": 8.174097664543526e-06,
+      "loss": 0.6159,
+      "step": 385
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.022417337666467,
+      "learning_rate": 8.195329087048833e-06,
+      "loss": 0.6367,
+      "step": 386
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.886779803282784,
+      "learning_rate": 8.21656050955414e-06,
+      "loss": 0.7315,
+      "step": 387
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6185410245742909,
+      "learning_rate": 8.237791932059449e-06,
+      "loss": 0.5285,
+      "step": 388
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6555034141382942,
+      "learning_rate": 8.259023354564757e-06,
+      "loss": 0.5407,
+      "step": 389
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.927509309026582,
+      "learning_rate": 8.280254777070064e-06,
+      "loss": 0.6634,
+      "step": 390
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8900249257788111,
+      "learning_rate": 8.301486199575372e-06,
+      "loss": 0.6216,
+      "step": 391
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8231308283868264,
+      "learning_rate": 8.32271762208068e-06,
+      "loss": 0.6526,
+      "step": 392
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7226958174541385,
+      "learning_rate": 8.343949044585989e-06,
+      "loss": 0.6514,
+      "step": 393
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6441010227146757,
+      "learning_rate": 8.365180467091297e-06,
+      "loss": 0.6123,
+      "step": 394
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7184500228931993,
+      "learning_rate": 8.386411889596604e-06,
+      "loss": 0.6132,
+      "step": 395
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7611063791793172,
+      "learning_rate": 8.407643312101912e-06,
+      "loss": 0.6107,
+      "step": 396
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8854254361174873,
+      "learning_rate": 8.42887473460722e-06,
+      "loss": 0.6482,
+      "step": 397
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7142906085084705,
+      "learning_rate": 8.450106157112528e-06,
+      "loss": 0.5991,
+      "step": 398
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7452467444618182,
+      "learning_rate": 8.471337579617835e-06,
+      "loss": 0.5906,
+      "step": 399
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.88893485363304,
+      "learning_rate": 8.492569002123143e-06,
+      "loss": 0.6391,
+      "step": 400
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7797099055102723,
+      "learning_rate": 8.513800424628451e-06,
+      "loss": 0.6288,
+      "step": 401
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1476881331334823,
+      "learning_rate": 8.53503184713376e-06,
+      "loss": 0.6295,
+      "step": 402
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6454433387905704,
+      "learning_rate": 8.556263269639066e-06,
+      "loss": 0.5851,
+      "step": 403
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9202915443187007,
+      "learning_rate": 8.577494692144374e-06,
+      "loss": 0.6669,
+      "step": 404
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7896740571157151,
+      "learning_rate": 8.598726114649683e-06,
+      "loss": 0.6247,
+      "step": 405
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.650371840959321,
+      "learning_rate": 8.619957537154991e-06,
+      "loss": 0.5711,
+      "step": 406
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2264582688372754,
+      "learning_rate": 8.641188959660298e-06,
+      "loss": 0.6571,
+      "step": 407
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7511960497743809,
+      "learning_rate": 8.662420382165606e-06,
+      "loss": 0.5488,
+      "step": 408
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8883132225129745,
+      "learning_rate": 8.683651804670914e-06,
+      "loss": 0.6498,
+      "step": 409
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8711475329433065,
+      "learning_rate": 8.704883227176222e-06,
+      "loss": 0.633,
+      "step": 410
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6482577969800928,
+      "learning_rate": 8.726114649681529e-06,
+      "loss": 0.5435,
+      "step": 411
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.833716792141639,
+      "learning_rate": 8.747346072186837e-06,
+      "loss": 0.5639,
+      "step": 412
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8404773082943031,
+      "learning_rate": 8.768577494692145e-06,
+      "loss": 0.6407,
+      "step": 413
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.003279144187182,
+      "learning_rate": 8.789808917197454e-06,
+      "loss": 0.6892,
+      "step": 414
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.702753050046326,
+      "learning_rate": 8.81104033970276e-06,
+      "loss": 0.5622,
+      "step": 415
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7055678953855339,
+      "learning_rate": 8.832271762208069e-06,
+      "loss": 0.573,
+      "step": 416
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7195947029496707,
+      "learning_rate": 8.853503184713377e-06,
+      "loss": 0.61,
+      "step": 417
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7367726612748673,
+      "learning_rate": 8.874734607218685e-06,
+      "loss": 0.5678,
+      "step": 418
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7274607829878597,
+      "learning_rate": 8.895966029723993e-06,
+      "loss": 0.529,
+      "step": 419
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7830343262063172,
+      "learning_rate": 8.9171974522293e-06,
+      "loss": 0.6487,
+      "step": 420
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8795400014954516,
+      "learning_rate": 8.938428874734608e-06,
+      "loss": 0.6744,
+      "step": 421
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8163787645669263,
+      "learning_rate": 8.959660297239916e-06,
+      "loss": 0.6919,
+      "step": 422
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6131252077612136,
+      "learning_rate": 8.980891719745225e-06,
+      "loss": 0.539,
+      "step": 423
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6577080763662211,
+      "learning_rate": 9.002123142250531e-06,
+      "loss": 0.5362,
+      "step": 424
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6754781830302256,
+      "learning_rate": 9.02335456475584e-06,
+      "loss": 0.618,
+      "step": 425
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6180744659367413,
+      "learning_rate": 9.044585987261148e-06,
+      "loss": 0.5789,
+      "step": 426
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7656745343043879,
+      "learning_rate": 9.065817409766456e-06,
+      "loss": 0.58,
+      "step": 427
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8047908307660389,
+      "learning_rate": 9.087048832271763e-06,
+      "loss": 0.6189,
+      "step": 428
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1378676066470146,
+      "learning_rate": 9.10828025477707e-06,
+      "loss": 0.6134,
+      "step": 429
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7408459076539501,
+      "learning_rate": 9.129511677282379e-06,
+      "loss": 0.6018,
+      "step": 430
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8835243577619359,
+      "learning_rate": 9.150743099787687e-06,
+      "loss": 0.6269,
+      "step": 431
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6478236226856178,
+      "learning_rate": 9.171974522292994e-06,
+      "loss": 0.5543,
+      "step": 432
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8829938350860344,
+      "learning_rate": 9.193205944798302e-06,
+      "loss": 0.6712,
+      "step": 433
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8020701009095713,
+      "learning_rate": 9.21443736730361e-06,
+      "loss": 0.5648,
+      "step": 434
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8097751212747626,
+      "learning_rate": 9.235668789808919e-06,
+      "loss": 0.6388,
+      "step": 435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7264245763854907,
+      "learning_rate": 9.256900212314225e-06,
+      "loss": 0.5783,
+      "step": 436
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6658669676284141,
+      "learning_rate": 9.278131634819534e-06,
+      "loss": 0.5424,
+      "step": 437
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7293560413896242,
+      "learning_rate": 9.299363057324842e-06,
+      "loss": 0.5849,
+      "step": 438
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9165722911894904,
+      "learning_rate": 9.32059447983015e-06,
+      "loss": 0.6842,
+      "step": 439
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6672157004325562,
+      "learning_rate": 9.341825902335457e-06,
+      "loss": 0.612,
+      "step": 440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.515072527151968,
+      "learning_rate": 9.363057324840765e-06,
+      "loss": 0.5953,
+      "step": 441
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8466680398556095,
+      "learning_rate": 9.384288747346073e-06,
+      "loss": 0.6494,
+      "step": 442
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.76149805712936,
+      "learning_rate": 9.405520169851381e-06,
+      "loss": 0.6328,
+      "step": 443
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.667792369022887,
+      "learning_rate": 9.426751592356688e-06,
+      "loss": 0.5881,
+      "step": 444
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1000601752879977,
+      "learning_rate": 9.447983014861996e-06,
+      "loss": 0.5828,
+      "step": 445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7977541069806383,
+      "learning_rate": 9.469214437367304e-06,
+      "loss": 0.6257,
+      "step": 446
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7554239986005157,
+      "learning_rate": 9.490445859872613e-06,
+      "loss": 0.5868,
+      "step": 447
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6902154989570397,
+      "learning_rate": 9.511677282377921e-06,
+      "loss": 0.5741,
+      "step": 448
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.831785734821144,
+      "learning_rate": 9.532908704883228e-06,
+      "loss": 0.6632,
+      "step": 449
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6589235504467976,
+      "learning_rate": 9.554140127388536e-06,
+      "loss": 0.5661,
+      "step": 450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6117670949273154,
+      "learning_rate": 9.575371549893844e-06,
+      "loss": 0.5249,
+      "step": 451
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.704911300521356,
+      "learning_rate": 9.596602972399152e-06,
+      "loss": 0.5271,
+      "step": 452
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8596611607511858,
+      "learning_rate": 9.617834394904459e-06,
+      "loss": 0.6802,
+      "step": 453
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8965616025174299,
+      "learning_rate": 9.639065817409767e-06,
+      "loss": 0.6273,
+      "step": 454
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6109748675543724,
+      "learning_rate": 9.660297239915075e-06,
+      "loss": 0.6023,
+      "step": 455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7991002535309961,
+      "learning_rate": 9.681528662420384e-06,
+      "loss": 0.6369,
+      "step": 456
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7452014876614284,
+      "learning_rate": 9.70276008492569e-06,
+      "loss": 0.5368,
+      "step": 457
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7116054497263848,
+      "learning_rate": 9.723991507430999e-06,
+      "loss": 0.647,
+      "step": 458
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7733574624574389,
+      "learning_rate": 9.745222929936307e-06,
+      "loss": 0.5834,
+      "step": 459
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8673038237405521,
+      "learning_rate": 9.766454352441615e-06,
+      "loss": 0.6557,
+      "step": 460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9087978224410032,
+      "learning_rate": 9.787685774946922e-06,
+      "loss": 0.6369,
+      "step": 461
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7248864196673108,
+      "learning_rate": 9.80891719745223e-06,
+      "loss": 0.5998,
+      "step": 462
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.762858181858747,
+      "learning_rate": 9.830148619957538e-06,
+      "loss": 0.6096,
+      "step": 463
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8189284828454165,
+      "learning_rate": 9.851380042462846e-06,
+      "loss": 0.6268,
+      "step": 464
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2234999520903838,
+      "learning_rate": 9.872611464968153e-06,
+      "loss": 0.6039,
+      "step": 465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6611390379618824,
+      "learning_rate": 9.893842887473461e-06,
+      "loss": 0.5561,
+      "step": 466
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7861456650459644,
+      "learning_rate": 9.91507430997877e-06,
+      "loss": 0.6825,
+      "step": 467
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7232391983502742,
+      "learning_rate": 9.936305732484078e-06,
+      "loss": 0.5668,
+      "step": 468
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8245054175555814,
+      "learning_rate": 9.957537154989384e-06,
+      "loss": 0.6465,
+      "step": 469
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8159840462089913,
+      "learning_rate": 9.978768577494693e-06,
+      "loss": 0.623,
+      "step": 470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6739818639719175,
+      "learning_rate": 1e-05,
+      "loss": 0.5829,
+      "step": 471
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9169394853492169,
+      "learning_rate": 9.99999989358299e-06,
+      "loss": 0.6543,
+      "step": 472
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.732398277025336,
+      "learning_rate": 9.99999957433196e-06,
+      "loss": 0.604,
+      "step": 473
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7935352329835511,
+      "learning_rate": 9.999999042246928e-06,
+      "loss": 0.6036,
+      "step": 474
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7204735579547549,
+      "learning_rate": 9.999998297327913e-06,
+      "loss": 0.567,
+      "step": 475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9466278469784851,
+      "learning_rate": 9.999997339574949e-06,
+      "loss": 0.6903,
+      "step": 476
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7510980657849371,
+      "learning_rate": 9.999996168988075e-06,
+      "loss": 0.5798,
+      "step": 477
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9251013947903287,
+      "learning_rate": 9.999994785567344e-06,
+      "loss": 0.6208,
+      "step": 478
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8823297094976791,
+      "learning_rate": 9.99999318931281e-06,
+      "loss": 0.6067,
+      "step": 479
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8889253143015425,
+      "learning_rate": 9.999991380224545e-06,
+      "loss": 0.6672,
+      "step": 480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8224626528968296,
+      "learning_rate": 9.999989358302623e-06,
+      "loss": 0.6605,
+      "step": 481
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5962750987251436,
+      "learning_rate": 9.999987123547133e-06,
+      "loss": 0.5454,
+      "step": 482
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8329766576540251,
+      "learning_rate": 9.999984675958169e-06,
+      "loss": 0.6366,
+      "step": 483
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8074101110714306,
+      "learning_rate": 9.999982015535834e-06,
+      "loss": 0.6583,
+      "step": 484
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6613443421470934,
+      "learning_rate": 9.999979142280246e-06,
+      "loss": 0.6069,
+      "step": 485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7004874091363438,
+      "learning_rate": 9.999976056191519e-06,
+      "loss": 0.5727,
+      "step": 486
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8654496022656083,
+      "learning_rate": 9.999972757269792e-06,
+      "loss": 0.6838,
+      "step": 487
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.723092832411239,
+      "learning_rate": 9.9999692455152e-06,
+      "loss": 0.5972,
+      "step": 488
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.718562574605943,
+      "learning_rate": 9.999965520927895e-06,
+      "loss": 0.5875,
+      "step": 489
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6778690028989336,
+      "learning_rate": 9.999961583508035e-06,
+      "loss": 0.5754,
+      "step": 490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8718546884186662,
+      "learning_rate": 9.99995743325579e-06,
+      "loss": 0.6625,
+      "step": 491
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7923702690792669,
+      "learning_rate": 9.999953070171334e-06,
+      "loss": 0.6471,
+      "step": 492
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7507109521830716,
+      "learning_rate": 9.999948494254853e-06,
+      "loss": 0.63,
+      "step": 493
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7205764419308681,
+      "learning_rate": 9.999943705506544e-06,
+      "loss": 0.5992,
+      "step": 494
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6601109934385242,
+      "learning_rate": 9.999938703926607e-06,
+      "loss": 0.528,
+      "step": 495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6786762625720932,
+      "learning_rate": 9.999933489515257e-06,
+      "loss": 0.5651,
+      "step": 496
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7622149354723696,
+      "learning_rate": 9.999928062272714e-06,
+      "loss": 0.6337,
+      "step": 497
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9885600548836091,
+      "learning_rate": 9.999922422199213e-06,
+      "loss": 0.6804,
+      "step": 498
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6422403188821002,
+      "learning_rate": 9.999916569294994e-06,
+      "loss": 0.5432,
+      "step": 499
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7370948573188223,
+      "learning_rate": 9.9999105035603e-06,
+      "loss": 0.5625,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7935001229962054,
+      "learning_rate": 9.999904224995397e-06,
+      "loss": 0.594,
+      "step": 501
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8847636308919531,
+      "learning_rate": 9.999897733600545e-06,
+      "loss": 0.6236,
+      "step": 502
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6425009679367006,
+      "learning_rate": 9.999891029376025e-06,
+      "loss": 0.6326,
+      "step": 503
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8989107169408984,
+      "learning_rate": 9.999884112322122e-06,
+      "loss": 0.6011,
+      "step": 504
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8916440421695172,
+      "learning_rate": 9.999876982439131e-06,
+      "loss": 0.698,
+      "step": 505
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8724879286301799,
+      "learning_rate": 9.999869639727353e-06,
+      "loss": 0.6846,
+      "step": 506
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7649817263081151,
+      "learning_rate": 9.999862084187101e-06,
+      "loss": 0.5864,
+      "step": 507
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.718616029914458,
+      "learning_rate": 9.999854315818697e-06,
+      "loss": 0.5718,
+      "step": 508
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8222526275223059,
+      "learning_rate": 9.999846334622474e-06,
+      "loss": 0.5788,
+      "step": 509
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.130838876693734,
+      "learning_rate": 9.999838140598768e-06,
+      "loss": 0.6831,
+      "step": 510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6932187212337148,
+      "learning_rate": 9.99982973374793e-06,
+      "loss": 0.5836,
+      "step": 511
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8214432283624149,
+      "learning_rate": 9.999821114070318e-06,
+      "loss": 0.6369,
+      "step": 512
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6525866559128067,
+      "learning_rate": 9.999812281566298e-06,
+      "loss": 0.5711,
+      "step": 513
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7239474453362629,
+      "learning_rate": 9.999803236236246e-06,
+      "loss": 0.5785,
+      "step": 514
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8555078103180713,
+      "learning_rate": 9.999793978080548e-06,
+      "loss": 0.6511,
+      "step": 515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.732197985942794,
+      "learning_rate": 9.999784507099598e-06,
+      "loss": 0.5622,
+      "step": 516
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8622453933353151,
+      "learning_rate": 9.999774823293794e-06,
+      "loss": 0.6547,
+      "step": 517
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9183637108699996,
+      "learning_rate": 9.999764926663558e-06,
+      "loss": 0.6939,
+      "step": 518
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8389181768232583,
+      "learning_rate": 9.999754817209305e-06,
+      "loss": 0.5699,
+      "step": 519
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6502745124123062,
+      "learning_rate": 9.999744494931465e-06,
+      "loss": 0.6191,
+      "step": 520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6886355269499652,
+      "learning_rate": 9.99973395983048e-06,
+      "loss": 0.5393,
+      "step": 521
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6612716942626704,
+      "learning_rate": 9.999723211906796e-06,
+      "loss": 0.6018,
+      "step": 522
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6501186337772761,
+      "learning_rate": 9.999712251160871e-06,
+      "loss": 0.5573,
+      "step": 523
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7706352731103632,
+      "learning_rate": 9.999701077593174e-06,
+      "loss": 0.6581,
+      "step": 524
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7730012720034379,
+      "learning_rate": 9.999689691204179e-06,
+      "loss": 0.5962,
+      "step": 525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9508155364434,
+      "learning_rate": 9.99967809199437e-06,
+      "loss": 0.6623,
+      "step": 526
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8060102874185446,
+      "learning_rate": 9.999666279964242e-06,
+      "loss": 0.6761,
+      "step": 527
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7066786391666761,
+      "learning_rate": 9.999654255114295e-06,
+      "loss": 0.5771,
+      "step": 528
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6828924290413217,
+      "learning_rate": 9.999642017445045e-06,
+      "loss": 0.62,
+      "step": 529
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7222614445163584,
+      "learning_rate": 9.99962956695701e-06,
+      "loss": 0.6126,
+      "step": 530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9046138767923666,
+      "learning_rate": 9.999616903650722e-06,
+      "loss": 0.5865,
+      "step": 531
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7886204704955198,
+      "learning_rate": 9.999604027526717e-06,
+      "loss": 0.65,
+      "step": 532
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0075595298045992,
+      "learning_rate": 9.999590938585546e-06,
+      "loss": 0.69,
+      "step": 533
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8124456029009315,
+      "learning_rate": 9.999577636827766e-06,
+      "loss": 0.6025,
+      "step": 534
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6329892453569808,
+      "learning_rate": 9.99956412225394e-06,
+      "loss": 0.5329,
+      "step": 535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7952247674357384,
+      "learning_rate": 9.999550394864647e-06,
+      "loss": 0.5825,
+      "step": 536
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8465928631531884,
+      "learning_rate": 9.99953645466047e-06,
+      "loss": 0.6563,
+      "step": 537
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8713340741755371,
+      "learning_rate": 9.999522301642004e-06,
+      "loss": 0.6791,
+      "step": 538
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9350628744751012,
+      "learning_rate": 9.999507935809848e-06,
+      "loss": 0.5694,
+      "step": 539
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7993387054659695,
+      "learning_rate": 9.999493357164616e-06,
+      "loss": 0.5733,
+      "step": 540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6539378896746085,
+      "learning_rate": 9.999478565706927e-06,
+      "loss": 0.5551,
+      "step": 541
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8243934943005577,
+      "learning_rate": 9.999463561437412e-06,
+      "loss": 0.624,
+      "step": 542
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7133367340956287,
+      "learning_rate": 9.999448344356709e-06,
+      "loss": 0.5759,
+      "step": 543
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6729087774343463,
+      "learning_rate": 9.999432914465466e-06,
+      "loss": 0.5409,
+      "step": 544
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.008624253443484,
+      "learning_rate": 9.99941727176434e-06,
+      "loss": 0.6462,
+      "step": 545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7191947866708757,
+      "learning_rate": 9.999401416253997e-06,
+      "loss": 0.5832,
+      "step": 546
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6854202092498776,
+      "learning_rate": 9.99938534793511e-06,
+      "loss": 0.5299,
+      "step": 547
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6434247509405175,
+      "learning_rate": 9.999369066808366e-06,
+      "loss": 0.5476,
+      "step": 548
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6140147352500417,
+      "learning_rate": 9.999352572874457e-06,
+      "loss": 0.5594,
+      "step": 549
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7705661429097387,
+      "learning_rate": 9.999335866134084e-06,
+      "loss": 0.6459,
+      "step": 550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8157044716675759,
+      "learning_rate": 9.999318946587957e-06,
+      "loss": 0.6638,
+      "step": 551
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5025190176934549,
+      "learning_rate": 9.9993018142368e-06,
+      "loss": 0.6143,
+      "step": 552
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6499419604536284,
+      "learning_rate": 9.999284469081338e-06,
+      "loss": 0.5732,
+      "step": 553
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7374540089225985,
+      "learning_rate": 9.999266911122314e-06,
+      "loss": 0.5951,
+      "step": 554
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6667930700380424,
+      "learning_rate": 9.999249140360473e-06,
+      "loss": 0.576,
+      "step": 555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0162824245358222,
+      "learning_rate": 9.99923115679657e-06,
+      "loss": 0.7426,
+      "step": 556
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6305009829157838,
+      "learning_rate": 9.999212960431372e-06,
+      "loss": 0.5965,
+      "step": 557
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.823049052290893,
+      "learning_rate": 9.999194551265653e-06,
+      "loss": 0.6681,
+      "step": 558
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7925088954710804,
+      "learning_rate": 9.9991759293002e-06,
+      "loss": 0.6394,
+      "step": 559
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6887087221153546,
+      "learning_rate": 9.9991570945358e-06,
+      "loss": 0.5601,
+      "step": 560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.631151150092096,
+      "learning_rate": 9.999138046973258e-06,
+      "loss": 0.5477,
+      "step": 561
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8605832852629615,
+      "learning_rate": 9.999118786613384e-06,
+      "loss": 0.6283,
+      "step": 562
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.739203631837093,
+      "learning_rate": 9.999099313457e-06,
+      "loss": 0.6127,
+      "step": 563
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6596350627733387,
+      "learning_rate": 9.999079627504931e-06,
+      "loss": 0.5736,
+      "step": 564
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4592063027815079,
+      "learning_rate": 9.999059728758018e-06,
+      "loss": 0.6749,
+      "step": 565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1222640617266986,
+      "learning_rate": 9.999039617217108e-06,
+      "loss": 0.7198,
+      "step": 566
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6920355737275525,
+      "learning_rate": 9.999019292883053e-06,
+      "loss": 0.612,
+      "step": 567
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.76675542132742,
+      "learning_rate": 9.998998755756724e-06,
+      "loss": 0.6004,
+      "step": 568
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8098835588375225,
+      "learning_rate": 9.998978005838993e-06,
+      "loss": 0.5792,
+      "step": 569
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9330560815304642,
+      "learning_rate": 9.99895704313074e-06,
+      "loss": 0.6745,
+      "step": 570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7107217808643186,
+      "learning_rate": 9.998935867632862e-06,
+      "loss": 0.5645,
+      "step": 571
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8628341720586082,
+      "learning_rate": 9.998914479346258e-06,
+      "loss": 0.5851,
+      "step": 572
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.812296064370831,
+      "learning_rate": 9.998892878271839e-06,
+      "loss": 0.6905,
+      "step": 573
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8344977204345659,
+      "learning_rate": 9.998871064410524e-06,
+      "loss": 0.636,
+      "step": 574
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6655544154860575,
+      "learning_rate": 9.998849037763243e-06,
+      "loss": 0.6042,
+      "step": 575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9039870258725992,
+      "learning_rate": 9.998826798330932e-06,
+      "loss": 0.7086,
+      "step": 576
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6918627084041002,
+      "learning_rate": 9.998804346114536e-06,
+      "loss": 0.532,
+      "step": 577
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6263729431215533,
+      "learning_rate": 9.998781681115017e-06,
+      "loss": 0.5717,
+      "step": 578
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9054569474491742,
+      "learning_rate": 9.998758803333333e-06,
+      "loss": 0.6493,
+      "step": 579
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7819365195426208,
+      "learning_rate": 9.99873571277046e-06,
+      "loss": 0.6549,
+      "step": 580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7845315517848321,
+      "learning_rate": 9.998712409427382e-06,
+      "loss": 0.6991,
+      "step": 581
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0283213258141806,
+      "learning_rate": 9.99868889330509e-06,
+      "loss": 0.669,
+      "step": 582
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9632780379974247,
+      "learning_rate": 9.998665164404586e-06,
+      "loss": 0.7075,
+      "step": 583
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7756967072338767,
+      "learning_rate": 9.99864122272688e-06,
+      "loss": 0.5049,
+      "step": 584
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9607889685736162,
+      "learning_rate": 9.99861706827299e-06,
+      "loss": 0.6408,
+      "step": 585
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6789333840195599,
+      "learning_rate": 9.998592701043945e-06,
+      "loss": 0.5208,
+      "step": 586
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7073847955888349,
+      "learning_rate": 9.998568121040781e-06,
+      "loss": 0.6186,
+      "step": 587
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8511169500473535,
+      "learning_rate": 9.998543328264544e-06,
+      "loss": 0.6493,
+      "step": 588
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6728585712796856,
+      "learning_rate": 9.998518322716294e-06,
+      "loss": 0.5862,
+      "step": 589
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8205208526183553,
+      "learning_rate": 9.99849310439709e-06,
+      "loss": 0.6667,
+      "step": 590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7534012734900305,
+      "learning_rate": 9.998467673308008e-06,
+      "loss": 0.6106,
+      "step": 591
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7119869181595472,
+      "learning_rate": 9.998442029450128e-06,
+      "loss": 0.5755,
+      "step": 592
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8701971631151063,
+      "learning_rate": 9.998416172824545e-06,
+      "loss": 0.7085,
+      "step": 593
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9227174594042213,
+      "learning_rate": 9.998390103432359e-06,
+      "loss": 0.6441,
+      "step": 594
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7216308708654693,
+      "learning_rate": 9.99836382127468e-06,
+      "loss": 0.5922,
+      "step": 595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5848478019701097,
+      "learning_rate": 9.998337326352623e-06,
+      "loss": 0.5544,
+      "step": 596
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.951394179802008,
+      "learning_rate": 9.99831061866732e-06,
+      "loss": 0.6307,
+      "step": 597
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6242384053692704,
+      "learning_rate": 9.998283698219905e-06,
+      "loss": 0.543,
+      "step": 598
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7794169851853867,
+      "learning_rate": 9.998256565011525e-06,
+      "loss": 0.6081,
+      "step": 599
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8420300848884748,
+      "learning_rate": 9.998229219043336e-06,
+      "loss": 0.5916,
+      "step": 600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.704643197328675,
+      "learning_rate": 9.9982016603165e-06,
+      "loss": 0.5532,
+      "step": 601
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6434413239654009,
+      "learning_rate": 9.998173888832193e-06,
+      "loss": 0.5382,
+      "step": 602
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.028914317228709,
+      "learning_rate": 9.998145904591595e-06,
+      "loss": 0.6082,
+      "step": 603
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6454334681598273,
+      "learning_rate": 9.998117707595898e-06,
+      "loss": 0.5557,
+      "step": 604
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6891323611593384,
+      "learning_rate": 9.9980892978463e-06,
+      "loss": 0.5964,
+      "step": 605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6483259854376953,
+      "learning_rate": 9.998060675344015e-06,
+      "loss": 0.5604,
+      "step": 606
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.640850410937195,
+      "learning_rate": 9.998031840090257e-06,
+      "loss": 0.5605,
+      "step": 607
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7237150931450929,
+      "learning_rate": 9.998002792086254e-06,
+      "loss": 0.5944,
+      "step": 608
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6996628810617805,
+      "learning_rate": 9.997973531333245e-06,
+      "loss": 0.5957,
+      "step": 609
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6173241916508667,
+      "learning_rate": 9.997944057832474e-06,
+      "loss": 0.5626,
+      "step": 610
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9164611669637491,
+      "learning_rate": 9.997914371585195e-06,
+      "loss": 0.677,
+      "step": 611
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7285589497635658,
+      "learning_rate": 9.997884472592672e-06,
+      "loss": 0.5631,
+      "step": 612
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9485477673636976,
+      "learning_rate": 9.997854360856178e-06,
+      "loss": 0.669,
+      "step": 613
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7781838708302896,
+      "learning_rate": 9.997824036376995e-06,
+      "loss": 0.6623,
+      "step": 614
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7898792801970184,
+      "learning_rate": 9.997793499156414e-06,
+      "loss": 0.6471,
+      "step": 615
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8308443987749227,
+      "learning_rate": 9.997762749195735e-06,
+      "loss": 0.6877,
+      "step": 616
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6310836629954972,
+      "learning_rate": 9.997731786496265e-06,
+      "loss": 0.5485,
+      "step": 617
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2053576611133063,
+      "learning_rate": 9.997700611059323e-06,
+      "loss": 0.6664,
+      "step": 618
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.613546971387751,
+      "learning_rate": 9.997669222886237e-06,
+      "loss": 0.5233,
+      "step": 619
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6745170096468901,
+      "learning_rate": 9.997637621978341e-06,
+      "loss": 0.5831,
+      "step": 620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8062537763927432,
+      "learning_rate": 9.997605808336985e-06,
+      "loss": 0.6216,
+      "step": 621
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6924571937105224,
+      "learning_rate": 9.997573781963517e-06,
+      "loss": 0.5789,
+      "step": 622
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.181097370897103,
+      "learning_rate": 9.997541542859305e-06,
+      "loss": 0.6104,
+      "step": 623
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5527977444336124,
+      "learning_rate": 9.997509091025718e-06,
+      "loss": 0.5088,
+      "step": 624
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8535412593908939,
+      "learning_rate": 9.99747642646414e-06,
+      "loss": 0.6658,
+      "step": 625
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8290599255212616,
+      "learning_rate": 9.997443549175957e-06,
+      "loss": 0.642,
+      "step": 626
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6585395907225443,
+      "learning_rate": 9.997410459162575e-06,
+      "loss": 0.5996,
+      "step": 627
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7643793451239477,
+      "learning_rate": 9.997377156425398e-06,
+      "loss": 0.6094,
+      "step": 628
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5850937661332478,
+      "learning_rate": 9.997343640965846e-06,
+      "loss": 0.5058,
+      "step": 629
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8798977741642525,
+      "learning_rate": 9.997309912785343e-06,
+      "loss": 0.6818,
+      "step": 630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7845680405955845,
+      "learning_rate": 9.997275971885326e-06,
+      "loss": 0.6399,
+      "step": 631
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.787892789994041,
+      "learning_rate": 9.997241818267241e-06,
+      "loss": 0.5836,
+      "step": 632
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.871851233272999,
+      "learning_rate": 9.997207451932539e-06,
+      "loss": 0.7112,
+      "step": 633
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8369811460124821,
+      "learning_rate": 9.997172872882685e-06,
+      "loss": 0.6088,
+      "step": 634
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7052094714802036,
+      "learning_rate": 9.997138081119153e-06,
+      "loss": 0.5751,
+      "step": 635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7405549824133622,
+      "learning_rate": 9.997103076643418e-06,
+      "loss": 0.5486,
+      "step": 636
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7189672087528193,
+      "learning_rate": 9.997067859456975e-06,
+      "loss": 0.5817,
+      "step": 637
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7107487540224503,
+      "learning_rate": 9.997032429561321e-06,
+      "loss": 0.5652,
+      "step": 638
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8769897123204996,
+      "learning_rate": 9.996996786957967e-06,
+      "loss": 0.6672,
+      "step": 639
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6401910642444841,
+      "learning_rate": 9.996960931648425e-06,
+      "loss": 0.5351,
+      "step": 640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6778235077811764,
+      "learning_rate": 9.996924863634225e-06,
+      "loss": 0.5921,
+      "step": 641
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7011876136636904,
+      "learning_rate": 9.996888582916902e-06,
+      "loss": 0.5723,
+      "step": 642
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0460066107180521,
+      "learning_rate": 9.996852089497999e-06,
+      "loss": 0.6511,
+      "step": 643
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6826216750177364,
+      "learning_rate": 9.996815383379072e-06,
+      "loss": 0.5215,
+      "step": 644
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6729038352173468,
+      "learning_rate": 9.99677846456168e-06,
+      "loss": 0.5201,
+      "step": 645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6388857321520266,
+      "learning_rate": 9.996741333047398e-06,
+      "loss": 0.5607,
+      "step": 646
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7085204374891683,
+      "learning_rate": 9.996703988837801e-06,
+      "loss": 0.5963,
+      "step": 647
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6648291904521154,
+      "learning_rate": 9.996666431934486e-06,
+      "loss": 0.5623,
+      "step": 648
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.791919546274351,
+      "learning_rate": 9.996628662339046e-06,
+      "loss": 0.6429,
+      "step": 649
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8183198332608097,
+      "learning_rate": 9.996590680053091e-06,
+      "loss": 0.632,
+      "step": 650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.806752433284419,
+      "learning_rate": 9.99655248507824e-06,
+      "loss": 0.6869,
+      "step": 651
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6296693161494359,
+      "learning_rate": 9.996514077416114e-06,
+      "loss": 0.5906,
+      "step": 652
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7090714175483955,
+      "learning_rate": 9.996475457068351e-06,
+      "loss": 0.5875,
+      "step": 653
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0149882615272856,
+      "learning_rate": 9.996436624036594e-06,
+      "loss": 0.6216,
+      "step": 654
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6379949441574454,
+      "learning_rate": 9.996397578322497e-06,
+      "loss": 0.5716,
+      "step": 655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8184009434954894,
+      "learning_rate": 9.996358319927719e-06,
+      "loss": 0.6285,
+      "step": 656
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7292977782606443,
+      "learning_rate": 9.996318848853936e-06,
+      "loss": 0.5799,
+      "step": 657
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6748165000254052,
+      "learning_rate": 9.996279165102824e-06,
+      "loss": 0.5708,
+      "step": 658
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8797755071234167,
+      "learning_rate": 9.996239268676075e-06,
+      "loss": 0.6819,
+      "step": 659
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7414200697991348,
+      "learning_rate": 9.996199159575385e-06,
+      "loss": 0.6368,
+      "step": 660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6793534011032392,
+      "learning_rate": 9.996158837802463e-06,
+      "loss": 0.6238,
+      "step": 661
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6365931996356293,
+      "learning_rate": 9.996118303359024e-06,
+      "loss": 0.6239,
+      "step": 662
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7063336894857222,
+      "learning_rate": 9.996077556246795e-06,
+      "loss": 0.7095,
+      "step": 663
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8983126405033272,
+      "learning_rate": 9.996036596467509e-06,
+      "loss": 0.6375,
+      "step": 664
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6875782930744684,
+      "learning_rate": 9.995995424022911e-06,
+      "loss": 0.5261,
+      "step": 665
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6755263244361941,
+      "learning_rate": 9.995954038914752e-06,
+      "loss": 0.5926,
+      "step": 666
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1520147850642788,
+      "learning_rate": 9.995912441144794e-06,
+      "loss": 0.7104,
+      "step": 667
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8267163628665787,
+      "learning_rate": 9.995870630714808e-06,
+      "loss": 0.6427,
+      "step": 668
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.839777535046645,
+      "learning_rate": 9.995828607626574e-06,
+      "loss": 0.6504,
+      "step": 669
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6164221903425307,
+      "learning_rate": 9.995786371881882e-06,
+      "loss": 0.5905,
+      "step": 670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9366572634947168,
+      "learning_rate": 9.995743923482527e-06,
+      "loss": 0.6999,
+      "step": 671
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7084248721753169,
+      "learning_rate": 9.995701262430317e-06,
+      "loss": 0.5855,
+      "step": 672
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8142888157647624,
+      "learning_rate": 9.995658388727067e-06,
+      "loss": 0.6274,
+      "step": 673
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.60095279503971,
+      "learning_rate": 9.995615302374607e-06,
+      "loss": 0.5436,
+      "step": 674
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6355827884716699,
+      "learning_rate": 9.995572003374765e-06,
+      "loss": 0.565,
+      "step": 675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6002682697245667,
+      "learning_rate": 9.995528491729386e-06,
+      "loss": 0.5332,
+      "step": 676
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7921462760949219,
+      "learning_rate": 9.995484767440321e-06,
+      "loss": 0.5832,
+      "step": 677
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8446136275255393,
+      "learning_rate": 9.995440830509437e-06,
+      "loss": 0.6251,
+      "step": 678
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.578197473122897,
+      "learning_rate": 9.995396680938599e-06,
+      "loss": 0.5717,
+      "step": 679
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5975808365786478,
+      "learning_rate": 9.995352318729685e-06,
+      "loss": 0.5526,
+      "step": 680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8402832168714275,
+      "learning_rate": 9.995307743884587e-06,
+      "loss": 0.6598,
+      "step": 681
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5955051072932334,
+      "learning_rate": 9.9952629564052e-06,
+      "loss": 0.5324,
+      "step": 682
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7020067178988065,
+      "learning_rate": 9.995217956293433e-06,
+      "loss": 0.5926,
+      "step": 683
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6494717514572398,
+      "learning_rate": 9.995172743551199e-06,
+      "loss": 0.5969,
+      "step": 684
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9143489878257325,
+      "learning_rate": 9.995127318180424e-06,
+      "loss": 0.6555,
+      "step": 685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.721704853088111,
+      "learning_rate": 9.99508168018304e-06,
+      "loss": 0.5903,
+      "step": 686
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6613346090827121,
+      "learning_rate": 9.995035829560993e-06,
+      "loss": 0.5517,
+      "step": 687
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5711868712156565,
+      "learning_rate": 9.994989766316232e-06,
+      "loss": 0.5386,
+      "step": 688
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6276912272892899,
+      "learning_rate": 9.994943490450719e-06,
+      "loss": 0.5614,
+      "step": 689
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6781272192810728,
+      "learning_rate": 9.99489700196642e-06,
+      "loss": 0.6198,
+      "step": 690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6205918684127034,
+      "learning_rate": 9.99485030086532e-06,
+      "loss": 0.621,
+      "step": 691
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0623697303149042,
+      "learning_rate": 9.994803387149403e-06,
+      "loss": 0.6651,
+      "step": 692
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6625208597496124,
+      "learning_rate": 9.994756260820668e-06,
+      "loss": 0.5595,
+      "step": 693
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.908463009621757,
+      "learning_rate": 9.994708921881118e-06,
+      "loss": 0.6057,
+      "step": 694
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7078131488575676,
+      "learning_rate": 9.994661370332772e-06,
+      "loss": 0.5556,
+      "step": 695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6595231667492926,
+      "learning_rate": 9.99461360617765e-06,
+      "loss": 0.538,
+      "step": 696
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6282873366296898,
+      "learning_rate": 9.99456562941779e-06,
+      "loss": 0.5978,
+      "step": 697
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8378046457361583,
+      "learning_rate": 9.99451744005523e-06,
+      "loss": 0.6169,
+      "step": 698
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7321909573353417,
+      "learning_rate": 9.994469038092021e-06,
+      "loss": 0.633,
+      "step": 699
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7138296810024128,
+      "learning_rate": 9.994420423530226e-06,
+      "loss": 0.5621,
+      "step": 700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8574973273647994,
+      "learning_rate": 9.994371596371916e-06,
+      "loss": 0.6913,
+      "step": 701
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7042105564808486,
+      "learning_rate": 9.994322556619165e-06,
+      "loss": 0.5727,
+      "step": 702
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8617735874889249,
+      "learning_rate": 9.994273304274063e-06,
+      "loss": 0.7115,
+      "step": 703
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0008493088743105,
+      "learning_rate": 9.994223839338704e-06,
+      "loss": 0.6561,
+      "step": 704
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7520301080440177,
+      "learning_rate": 9.994174161815198e-06,
+      "loss": 0.6137,
+      "step": 705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6516214975658886,
+      "learning_rate": 9.994124271705654e-06,
+      "loss": 0.5581,
+      "step": 706
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9097192762638664,
+      "learning_rate": 9.994074169012201e-06,
+      "loss": 0.5854,
+      "step": 707
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8444910501822214,
+      "learning_rate": 9.99402385373697e-06,
+      "loss": 0.5931,
+      "step": 708
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6834082983943086,
+      "learning_rate": 9.9939733258821e-06,
+      "loss": 0.5772,
+      "step": 709
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9681715674561516,
+      "learning_rate": 9.993922585449745e-06,
+      "loss": 0.6371,
+      "step": 710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9051820415048046,
+      "learning_rate": 9.993871632442065e-06,
+      "loss": 0.6557,
+      "step": 711
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7707641503024311,
+      "learning_rate": 9.993820466861225e-06,
+      "loss": 0.6534,
+      "step": 712
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2328567189492285,
+      "learning_rate": 9.99376908870941e-06,
+      "loss": 0.6417,
+      "step": 713
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7516038633024393,
+      "learning_rate": 9.993717497988797e-06,
+      "loss": 0.6337,
+      "step": 714
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7045844606273769,
+      "learning_rate": 9.993665694701591e-06,
+      "loss": 0.5938,
+      "step": 715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6691251105720674,
+      "learning_rate": 9.993613678849994e-06,
+      "loss": 0.5311,
+      "step": 716
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8419025097296773,
+      "learning_rate": 9.99356145043622e-06,
+      "loss": 0.6553,
+      "step": 717
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6814007274604756,
+      "learning_rate": 9.993509009462491e-06,
+      "loss": 0.5697,
+      "step": 718
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5878337898826469,
+      "learning_rate": 9.993456355931042e-06,
+      "loss": 0.5071,
+      "step": 719
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7779635187798183,
+      "learning_rate": 9.993403489844112e-06,
+      "loss": 0.6374,
+      "step": 720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8100625867571185,
+      "learning_rate": 9.99335041120395e-06,
+      "loss": 0.5503,
+      "step": 721
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6768655524400484,
+      "learning_rate": 9.99329712001282e-06,
+      "loss": 0.5852,
+      "step": 722
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7387183957894116,
+      "learning_rate": 9.993243616272987e-06,
+      "loss": 0.6057,
+      "step": 723
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6497037461439694,
+      "learning_rate": 9.993189899986731e-06,
+      "loss": 0.5253,
+      "step": 724
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8513723100526019,
+      "learning_rate": 9.993135971156335e-06,
+      "loss": 0.6671,
+      "step": 725
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8503166815417971,
+      "learning_rate": 9.993081829784098e-06,
+      "loss": 0.7132,
+      "step": 726
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7190913432055303,
+      "learning_rate": 9.993027475872322e-06,
+      "loss": 0.5251,
+      "step": 727
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6644874297155562,
+      "learning_rate": 9.992972909423321e-06,
+      "loss": 0.5253,
+      "step": 728
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8234238920892054,
+      "learning_rate": 9.99291813043942e-06,
+      "loss": 0.5972,
+      "step": 729
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7352248725035845,
+      "learning_rate": 9.992863138922949e-06,
+      "loss": 0.6274,
+      "step": 730
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7644661183425183,
+      "learning_rate": 9.992807934876248e-06,
+      "loss": 0.685,
+      "step": 731
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7556809583325526,
+      "learning_rate": 9.99275251830167e-06,
+      "loss": 0.5298,
+      "step": 732
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.824620178065385,
+      "learning_rate": 9.99269688920157e-06,
+      "loss": 0.6469,
+      "step": 733
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7232993765683309,
+      "learning_rate": 9.992641047578319e-06,
+      "loss": 0.5896,
+      "step": 734
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6835725439560649,
+      "learning_rate": 9.99258499343429e-06,
+      "loss": 0.5439,
+      "step": 735
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6363589148560473,
+      "learning_rate": 9.992528726771875e-06,
+      "loss": 0.5456,
+      "step": 736
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.656622508428008,
+      "learning_rate": 9.992472247593466e-06,
+      "loss": 0.5427,
+      "step": 737
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3032360495897406,
+      "learning_rate": 9.992415555901466e-06,
+      "loss": 0.648,
+      "step": 738
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7557967211581249,
+      "learning_rate": 9.99235865169829e-06,
+      "loss": 0.627,
+      "step": 739
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6555105421279371,
+      "learning_rate": 9.992301534986359e-06,
+      "loss": 0.5434,
+      "step": 740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7438722139028712,
+      "learning_rate": 9.992244205768104e-06,
+      "loss": 0.5501,
+      "step": 741
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7776058377148094,
+      "learning_rate": 9.992186664045966e-06,
+      "loss": 0.5536,
+      "step": 742
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6041178167844797,
+      "learning_rate": 9.992128909822395e-06,
+      "loss": 0.5535,
+      "step": 743
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6041064112548524,
+      "learning_rate": 9.99207094309985e-06,
+      "loss": 0.4954,
+      "step": 744
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6782353751153032,
+      "learning_rate": 9.992012763880797e-06,
+      "loss": 0.6132,
+      "step": 745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5807807179807045,
+      "learning_rate": 9.991954372167711e-06,
+      "loss": 0.4822,
+      "step": 746
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8909082671096811,
+      "learning_rate": 9.991895767963082e-06,
+      "loss": 0.6357,
+      "step": 747
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6784308680180855,
+      "learning_rate": 9.991836951269401e-06,
+      "loss": 0.5967,
+      "step": 748
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8851869588465735,
+      "learning_rate": 9.99177792208917e-06,
+      "loss": 0.6157,
+      "step": 749
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.80918196157227,
+      "learning_rate": 9.991718680424906e-06,
+      "loss": 0.6044,
+      "step": 750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7570481980607737,
+      "learning_rate": 9.991659226279132e-06,
+      "loss": 0.6415,
+      "step": 751
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6209879202808876,
+      "learning_rate": 9.991599559654372e-06,
+      "loss": 0.5686,
+      "step": 752
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5636213452795127,
+      "learning_rate": 9.991539680553171e-06,
+      "loss": 0.5461,
+      "step": 753
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7768031019190919,
+      "learning_rate": 9.991479588978077e-06,
+      "loss": 0.6046,
+      "step": 754
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8738921674332195,
+      "learning_rate": 9.991419284931645e-06,
+      "loss": 0.6363,
+      "step": 755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9867078491125648,
+      "learning_rate": 9.991358768416449e-06,
+      "loss": 0.5987,
+      "step": 756
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6157519853697115,
+      "learning_rate": 9.991298039435055e-06,
+      "loss": 0.5974,
+      "step": 757
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8817679833279981,
+      "learning_rate": 9.991237097990057e-06,
+      "loss": 0.5939,
+      "step": 758
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6540140299071304,
+      "learning_rate": 9.991175944084044e-06,
+      "loss": 0.54,
+      "step": 759
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7985386733836422,
+      "learning_rate": 9.991114577719621e-06,
+      "loss": 0.6122,
+      "step": 760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.664628522098663,
+      "learning_rate": 9.9910529988994e-06,
+      "loss": 0.5446,
+      "step": 761
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7417870718986781,
+      "learning_rate": 9.990991207626e-06,
+      "loss": 0.6229,
+      "step": 762
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6035253140147998,
+      "learning_rate": 9.990929203902056e-06,
+      "loss": 0.5411,
+      "step": 763
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7360947084341074,
+      "learning_rate": 9.990866987730204e-06,
+      "loss": 0.5483,
+      "step": 764
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6659011890744978,
+      "learning_rate": 9.990804559113093e-06,
+      "loss": 0.509,
+      "step": 765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7637223777381329,
+      "learning_rate": 9.990741918053376e-06,
+      "loss": 0.601,
+      "step": 766
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.817278269228528,
+      "learning_rate": 9.99067906455373e-06,
+      "loss": 0.6557,
+      "step": 767
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6655728021288501,
+      "learning_rate": 9.99061599861682e-06,
+      "loss": 0.5226,
+      "step": 768
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9914790343979668,
+      "learning_rate": 9.990552720245336e-06,
+      "loss": 0.6255,
+      "step": 769
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7603987378373237,
+      "learning_rate": 9.99048922944197e-06,
+      "loss": 0.5869,
+      "step": 770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8781397184181137,
+      "learning_rate": 9.990425526209424e-06,
+      "loss": 0.6306,
+      "step": 771
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7860454394941983,
+      "learning_rate": 9.990361610550412e-06,
+      "loss": 0.6214,
+      "step": 772
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6146630025732014,
+      "learning_rate": 9.990297482467653e-06,
+      "loss": 0.582,
+      "step": 773
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6136480741260867,
+      "learning_rate": 9.990233141963877e-06,
+      "loss": 0.5521,
+      "step": 774
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8202346777365337,
+      "learning_rate": 9.990168589041821e-06,
+      "loss": 0.6426,
+      "step": 775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8162614768870611,
+      "learning_rate": 9.990103823704236e-06,
+      "loss": 0.6168,
+      "step": 776
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.822384912582235,
+      "learning_rate": 9.990038845953876e-06,
+      "loss": 0.6485,
+      "step": 777
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6787019985342974,
+      "learning_rate": 9.98997365579351e-06,
+      "loss": 0.5608,
+      "step": 778
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6934141057209189,
+      "learning_rate": 9.989908253225911e-06,
+      "loss": 0.5967,
+      "step": 779
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.860039086305572,
+      "learning_rate": 9.989842638253861e-06,
+      "loss": 0.7043,
+      "step": 780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6735360534206339,
+      "learning_rate": 9.989776810880156e-06,
+      "loss": 0.5952,
+      "step": 781
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6717695884994594,
+      "learning_rate": 9.9897107711076e-06,
+      "loss": 0.5797,
+      "step": 782
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6516625962471542,
+      "learning_rate": 9.989644518938998e-06,
+      "loss": 0.5605,
+      "step": 783
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9424134113324836,
+      "learning_rate": 9.989578054377174e-06,
+      "loss": 0.6219,
+      "step": 784
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7819554334442894,
+      "learning_rate": 9.989511377424957e-06,
+      "loss": 0.5281,
+      "step": 785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8878934838991167,
+      "learning_rate": 9.989444488085185e-06,
+      "loss": 0.6686,
+      "step": 786
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5945890322248437,
+      "learning_rate": 9.989377386360706e-06,
+      "loss": 0.5234,
+      "step": 787
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7394100402484586,
+      "learning_rate": 9.989310072254375e-06,
+      "loss": 0.5295,
+      "step": 788
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6550729403258995,
+      "learning_rate": 9.989242545769056e-06,
+      "loss": 0.5895,
+      "step": 789
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8327940750285137,
+      "learning_rate": 9.989174806907627e-06,
+      "loss": 0.672,
+      "step": 790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9044076708722738,
+      "learning_rate": 9.989106855672968e-06,
+      "loss": 0.6254,
+      "step": 791
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.840277815762319,
+      "learning_rate": 9.989038692067974e-06,
+      "loss": 0.6586,
+      "step": 792
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6758925195276294,
+      "learning_rate": 9.988970316095547e-06,
+      "loss": 0.5875,
+      "step": 793
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8431847504791554,
+      "learning_rate": 9.988901727758594e-06,
+      "loss": 0.6252,
+      "step": 794
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.227123737024226,
+      "learning_rate": 9.988832927060038e-06,
+      "loss": 0.741,
+      "step": 795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8854218475472951,
+      "learning_rate": 9.988763914002806e-06,
+      "loss": 0.6091,
+      "step": 796
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5903794308945736,
+      "learning_rate": 9.988694688589836e-06,
+      "loss": 0.6588,
+      "step": 797
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7046035355578506,
+      "learning_rate": 9.988625250824074e-06,
+      "loss": 0.5571,
+      "step": 798
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.639894445052264,
+      "learning_rate": 9.988555600708476e-06,
+      "loss": 0.5296,
+      "step": 799
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6306065126784005,
+      "learning_rate": 9.98848573824601e-06,
+      "loss": 0.5398,
+      "step": 800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8263720474269912,
+      "learning_rate": 9.988415663439645e-06,
+      "loss": 0.6316,
+      "step": 801
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6595772593698904,
+      "learning_rate": 9.988345376292366e-06,
+      "loss": 0.4882,
+      "step": 802
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9817781299679428,
+      "learning_rate": 9.988274876807164e-06,
+      "loss": 0.6646,
+      "step": 803
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8816016131007491,
+      "learning_rate": 9.988204164987042e-06,
+      "loss": 0.6662,
+      "step": 804
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6877624983632619,
+      "learning_rate": 9.988133240835008e-06,
+      "loss": 0.5817,
+      "step": 805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1979913318985735,
+      "learning_rate": 9.988062104354083e-06,
+      "loss": 0.6077,
+      "step": 806
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.818338647993114,
+      "learning_rate": 9.987990755547293e-06,
+      "loss": 0.662,
+      "step": 807
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6286575467844161,
+      "learning_rate": 9.987919194417675e-06,
+      "loss": 0.5648,
+      "step": 808
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.668826912313625,
+      "learning_rate": 9.987847420968278e-06,
+      "loss": 0.5823,
+      "step": 809
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8283074214210586,
+      "learning_rate": 9.987775435202153e-06,
+      "loss": 0.6711,
+      "step": 810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5467020703048837,
+      "learning_rate": 9.987703237122366e-06,
+      "loss": 0.515,
+      "step": 811
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.991593659635225,
+      "learning_rate": 9.987630826731993e-06,
+      "loss": 0.6609,
+      "step": 812
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7969645835899805,
+      "learning_rate": 9.987558204034114e-06,
+      "loss": 0.6772,
+      "step": 813
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8937788622731796,
+      "learning_rate": 9.987485369031817e-06,
+      "loss": 0.7098,
+      "step": 814
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.608424689955248,
+      "learning_rate": 9.987412321728209e-06,
+      "loss": 0.5166,
+      "step": 815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8267505483392998,
+      "learning_rate": 9.987339062126394e-06,
+      "loss": 0.6073,
+      "step": 816
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6303518988797593,
+      "learning_rate": 9.987265590229494e-06,
+      "loss": 0.5204,
+      "step": 817
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6519974039844972,
+      "learning_rate": 9.987191906040634e-06,
+      "loss": 0.5612,
+      "step": 818
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6690501483398921,
+      "learning_rate": 9.98711800956295e-06,
+      "loss": 0.544,
+      "step": 819
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6613313944575195,
+      "learning_rate": 9.98704390079959e-06,
+      "loss": 0.5547,
+      "step": 820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2262546884070182,
+      "learning_rate": 9.986969579753706e-06,
+      "loss": 0.6491,
+      "step": 821
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6410866164962201,
+      "learning_rate": 9.986895046428467e-06,
+      "loss": 0.5322,
+      "step": 822
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7479005045701804,
+      "learning_rate": 9.986820300827038e-06,
+      "loss": 0.5812,
+      "step": 823
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8574836652317681,
+      "learning_rate": 9.986745342952605e-06,
+      "loss": 0.6477,
+      "step": 824
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.808972861030263,
+      "learning_rate": 9.986670172808359e-06,
+      "loss": 0.6282,
+      "step": 825
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7115535529703887,
+      "learning_rate": 9.9865947903975e-06,
+      "loss": 0.499,
+      "step": 826
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7978214995594192,
+      "learning_rate": 9.986519195723233e-06,
+      "loss": 0.649,
+      "step": 827
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6958941154244207,
+      "learning_rate": 9.986443388788781e-06,
+      "loss": 0.5564,
+      "step": 828
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8236487676418658,
+      "learning_rate": 9.986367369597366e-06,
+      "loss": 0.5772,
+      "step": 829
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7713431504854724,
+      "learning_rate": 9.986291138152227e-06,
+      "loss": 0.572,
+      "step": 830
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6744967713967663,
+      "learning_rate": 9.986214694456609e-06,
+      "loss": 0.5558,
+      "step": 831
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7719304482937637,
+      "learning_rate": 9.986138038513765e-06,
+      "loss": 0.596,
+      "step": 832
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6605250408593111,
+      "learning_rate": 9.986061170326958e-06,
+      "loss": 0.6089,
+      "step": 833
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7431533406246252,
+      "learning_rate": 9.98598408989946e-06,
+      "loss": 0.544,
+      "step": 834
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6748294392612384,
+      "learning_rate": 9.98590679723455e-06,
+      "loss": 0.5397,
+      "step": 835
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7118939115746294,
+      "learning_rate": 9.985829292335523e-06,
+      "loss": 0.6153,
+      "step": 836
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6076472785108435,
+      "learning_rate": 9.985751575205676e-06,
+      "loss": 0.5244,
+      "step": 837
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8751842414285672,
+      "learning_rate": 9.985673645848315e-06,
+      "loss": 0.6215,
+      "step": 838
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8569561121943189,
+      "learning_rate": 9.985595504266758e-06,
+      "loss": 0.609,
+      "step": 839
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8646887324930379,
+      "learning_rate": 9.985517150464335e-06,
+      "loss": 0.7117,
+      "step": 840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7605808467254423,
+      "learning_rate": 9.985438584444375e-06,
+      "loss": 0.5617,
+      "step": 841
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7730492847886752,
+      "learning_rate": 9.985359806210229e-06,
+      "loss": 0.6153,
+      "step": 842
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6678351603223434,
+      "learning_rate": 9.985280815765244e-06,
+      "loss": 0.5506,
+      "step": 843
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6927775493943853,
+      "learning_rate": 9.985201613112788e-06,
+      "loss": 0.5791,
+      "step": 844
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8481325219175609,
+      "learning_rate": 9.985122198256227e-06,
+      "loss": 0.5568,
+      "step": 845
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5960932300090971,
+      "learning_rate": 9.985042571198947e-06,
+      "loss": 0.5035,
+      "step": 846
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.875815844031176,
+      "learning_rate": 9.984962731944332e-06,
+      "loss": 0.659,
+      "step": 847
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7295163582351202,
+      "learning_rate": 9.984882680495784e-06,
+      "loss": 0.6118,
+      "step": 848
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7520869448489738,
+      "learning_rate": 9.984802416856711e-06,
+      "loss": 0.5664,
+      "step": 849
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8502409940694009,
+      "learning_rate": 9.984721941030528e-06,
+      "loss": 0.6992,
+      "step": 850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7819413543841294,
+      "learning_rate": 9.984641253020659e-06,
+      "loss": 0.5361,
+      "step": 851
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.871924910992041,
+      "learning_rate": 9.984560352830542e-06,
+      "loss": 0.6563,
+      "step": 852
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7098584975906701,
+      "learning_rate": 9.98447924046362e-06,
+      "loss": 0.5937,
+      "step": 853
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6019504772808552,
+      "learning_rate": 9.984397915923344e-06,
+      "loss": 0.5419,
+      "step": 854
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8800667340065077,
+      "learning_rate": 9.984316379213175e-06,
+      "loss": 0.6453,
+      "step": 855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9498703164426573,
+      "learning_rate": 9.984234630336586e-06,
+      "loss": 0.6814,
+      "step": 856
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6556240441790118,
+      "learning_rate": 9.984152669297058e-06,
+      "loss": 0.5306,
+      "step": 857
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9087981299754682,
+      "learning_rate": 9.984070496098076e-06,
+      "loss": 0.6071,
+      "step": 858
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5963703935302598,
+      "learning_rate": 9.983988110743141e-06,
+      "loss": 0.5068,
+      "step": 859
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7657075869969863,
+      "learning_rate": 9.983905513235758e-06,
+      "loss": 0.5896,
+      "step": 860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7436250195216065,
+      "learning_rate": 9.983822703579445e-06,
+      "loss": 0.5575,
+      "step": 861
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6023354584527244,
+      "learning_rate": 9.983739681777723e-06,
+      "loss": 0.554,
+      "step": 862
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1627523945571974,
+      "learning_rate": 9.983656447834129e-06,
+      "loss": 0.6125,
+      "step": 863
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8267175821570756,
+      "learning_rate": 9.983573001752208e-06,
+      "loss": 0.6356,
+      "step": 864
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6092360522288158,
+      "learning_rate": 9.983489343535506e-06,
+      "loss": 0.5403,
+      "step": 865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.908618839583729,
+      "learning_rate": 9.98340547318759e-06,
+      "loss": 0.6543,
+      "step": 866
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.125831052765039,
+      "learning_rate": 9.983321390712028e-06,
+      "loss": 0.6917,
+      "step": 867
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7682568756384482,
+      "learning_rate": 9.983237096112397e-06,
+      "loss": 0.6885,
+      "step": 868
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6470432155642117,
+      "learning_rate": 9.98315258939229e-06,
+      "loss": 0.5309,
+      "step": 869
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.084456252229607,
+      "learning_rate": 9.983067870555297e-06,
+      "loss": 0.6374,
+      "step": 870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8018245820192976,
+      "learning_rate": 9.98298293960503e-06,
+      "loss": 0.5757,
+      "step": 871
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9643134256057387,
+      "learning_rate": 9.982897796545104e-06,
+      "loss": 0.6722,
+      "step": 872
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8580838189426214,
+      "learning_rate": 9.982812441379141e-06,
+      "loss": 0.6549,
+      "step": 873
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7800072711229105,
+      "learning_rate": 9.982726874110776e-06,
+      "loss": 0.687,
+      "step": 874
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0363934860983977,
+      "learning_rate": 9.982641094743648e-06,
+      "loss": 0.6925,
+      "step": 875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7987480249623445,
+      "learning_rate": 9.982555103281413e-06,
+      "loss": 0.6675,
+      "step": 876
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6689629806115877,
+      "learning_rate": 9.982468899727728e-06,
+      "loss": 0.5386,
+      "step": 877
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9533975614522542,
+      "learning_rate": 9.982382484086263e-06,
+      "loss": 0.6647,
+      "step": 878
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7022065148065463,
+      "learning_rate": 9.9822958563607e-06,
+      "loss": 0.5639,
+      "step": 879
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8822886302244307,
+      "learning_rate": 9.98220901655472e-06,
+      "loss": 0.6937,
+      "step": 880
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9593099526128606,
+      "learning_rate": 9.982121964672027e-06,
+      "loss": 0.6368,
+      "step": 881
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8376183826694517,
+      "learning_rate": 9.98203470071632e-06,
+      "loss": 0.5754,
+      "step": 882
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8580188257672448,
+      "learning_rate": 9.981947224691316e-06,
+      "loss": 0.6551,
+      "step": 883
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6865148549538115,
+      "learning_rate": 9.98185953660074e-06,
+      "loss": 0.6089,
+      "step": 884
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.749497520981886,
+      "learning_rate": 9.981771636448323e-06,
+      "loss": 0.5604,
+      "step": 885
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0383945444894498,
+      "learning_rate": 9.981683524237805e-06,
+      "loss": 0.7119,
+      "step": 886
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.736463033196381,
+      "learning_rate": 9.98159519997294e-06,
+      "loss": 0.5953,
+      "step": 887
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.618337131288357,
+      "learning_rate": 9.981506663657486e-06,
+      "loss": 0.5805,
+      "step": 888
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6831445190302305,
+      "learning_rate": 9.981417915295213e-06,
+      "loss": 0.5492,
+      "step": 889
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8456877619742489,
+      "learning_rate": 9.981328954889896e-06,
+      "loss": 0.6944,
+      "step": 890
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8339034668691195,
+      "learning_rate": 9.981239782445325e-06,
+      "loss": 0.6603,
+      "step": 891
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8107358052589377,
+      "learning_rate": 9.981150397965293e-06,
+      "loss": 0.6619,
+      "step": 892
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5933002724612979,
+      "learning_rate": 9.981060801453605e-06,
+      "loss": 0.5694,
+      "step": 893
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.703868613588881,
+      "learning_rate": 9.980970992914079e-06,
+      "loss": 0.5536,
+      "step": 894
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0813069844773486,
+      "learning_rate": 9.980880972350533e-06,
+      "loss": 0.6424,
+      "step": 895
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8435970727418484,
+      "learning_rate": 9.980790739766801e-06,
+      "loss": 0.6975,
+      "step": 896
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6400389588171608,
+      "learning_rate": 9.980700295166724e-06,
+      "loss": 0.5398,
+      "step": 897
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0833546007836785,
+      "learning_rate": 9.98060963855415e-06,
+      "loss": 0.7218,
+      "step": 898
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6181242714505477,
+      "learning_rate": 9.980518769932938e-06,
+      "loss": 0.5475,
+      "step": 899
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8244564941672826,
+      "learning_rate": 9.980427689306962e-06,
+      "loss": 0.6312,
+      "step": 900
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9318019554131991,
+      "learning_rate": 9.98033639668009e-06,
+      "loss": 0.6611,
+      "step": 901
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4396113769073684,
+      "learning_rate": 9.980244892056216e-06,
+      "loss": 0.711,
+      "step": 902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9624421894783178,
+      "learning_rate": 9.980153175439229e-06,
+      "loss": 0.6278,
+      "step": 903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6600662675780681,
+      "learning_rate": 9.980061246833037e-06,
+      "loss": 0.5735,
+      "step": 904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8117856760783336,
+      "learning_rate": 9.979969106241551e-06,
+      "loss": 0.6374,
+      "step": 905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6889655407137212,
+      "learning_rate": 9.979876753668695e-06,
+      "loss": 0.6147,
+      "step": 906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7440401528217445,
+      "learning_rate": 9.979784189118398e-06,
+      "loss": 0.5489,
+      "step": 907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7987482256491769,
+      "learning_rate": 9.979691412594601e-06,
+      "loss": 0.6265,
+      "step": 908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8769697592520569,
+      "learning_rate": 9.979598424101253e-06,
+      "loss": 0.6605,
+      "step": 909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5906352109625501,
+      "learning_rate": 9.979505223642314e-06,
+      "loss": 0.5676,
+      "step": 910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.811207726950151,
+      "learning_rate": 9.97941181122175e-06,
+      "loss": 0.6785,
+      "step": 911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8088554766434953,
+      "learning_rate": 9.979318186843536e-06,
+      "loss": 0.6877,
+      "step": 912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8193905623642039,
+      "learning_rate": 9.979224350511658e-06,
+      "loss": 0.6505,
+      "step": 913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7438195458920674,
+      "learning_rate": 9.979130302230112e-06,
+      "loss": 0.5947,
+      "step": 914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8952939406382615,
+      "learning_rate": 9.979036042002899e-06,
+      "loss": 0.7302,
+      "step": 915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6503815007052984,
+      "learning_rate": 9.978941569834033e-06,
+      "loss": 0.574,
+      "step": 916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6709939905326485,
+      "learning_rate": 9.978846885727536e-06,
+      "loss": 0.5437,
+      "step": 917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8905563115820866,
+      "learning_rate": 9.978751989687437e-06,
+      "loss": 0.6812,
+      "step": 918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.844480162747943,
+      "learning_rate": 9.978656881717774e-06,
+      "loss": 0.6143,
+      "step": 919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6506977904792782,
+      "learning_rate": 9.978561561822598e-06,
+      "loss": 0.5209,
+      "step": 920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7591810938046205,
+      "learning_rate": 9.978466030005964e-06,
+      "loss": 0.5882,
+      "step": 921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.789580194552697,
+      "learning_rate": 9.97837028627194e-06,
+      "loss": 0.6269,
+      "step": 922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.61162725932201,
+      "learning_rate": 9.978274330624604e-06,
+      "loss": 0.5384,
+      "step": 923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7958299395610733,
+      "learning_rate": 9.978178163068035e-06,
+      "loss": 0.6472,
+      "step": 924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.718883200546543,
+      "learning_rate": 9.978081783606332e-06,
+      "loss": 0.5932,
+      "step": 925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0171229312281451,
+      "learning_rate": 9.977985192243596e-06,
+      "loss": 0.6725,
+      "step": 926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9366746409763742,
+      "learning_rate": 9.977888388983935e-06,
+      "loss": 0.7197,
+      "step": 927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6951243322287972,
+      "learning_rate": 9.977791373831474e-06,
+      "loss": 0.5666,
+      "step": 928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9492044887336627,
+      "learning_rate": 9.97769414679034e-06,
+      "loss": 0.6385,
+      "step": 929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6958637272327954,
+      "learning_rate": 9.977596707864673e-06,
+      "loss": 0.569,
+      "step": 930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6150529362764223,
+      "learning_rate": 9.97749905705862e-06,
+      "loss": 0.5312,
+      "step": 931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8477634045500256,
+      "learning_rate": 9.97740119437634e-06,
+      "loss": 0.6062,
+      "step": 932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8303457600093067,
+      "learning_rate": 9.977303119821994e-06,
+      "loss": 0.6604,
+      "step": 933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6807067924550453,
+      "learning_rate": 9.977204833399761e-06,
+      "loss": 0.5853,
+      "step": 934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7044752600345968,
+      "learning_rate": 9.977106335113821e-06,
+      "loss": 0.5931,
+      "step": 935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6435129007633353,
+      "learning_rate": 9.977007624968371e-06,
+      "loss": 0.5699,
+      "step": 936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6334797029424789,
+      "learning_rate": 9.97690870296761e-06,
+      "loss": 0.5531,
+      "step": 937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6622043096372564,
+      "learning_rate": 9.976809569115749e-06,
+      "loss": 0.6299,
+      "step": 938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6378979654875966,
+      "learning_rate": 9.97671022341701e-06,
+      "loss": 0.5171,
+      "step": 939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5915013537029304,
+      "learning_rate": 9.976610665875616e-06,
+      "loss": 0.5631,
+      "step": 940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6343711762016183,
+      "learning_rate": 9.976510896495813e-06,
+      "loss": 0.5886,
+      "step": 941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7789674759810145,
+      "learning_rate": 9.976410915281842e-06,
+      "loss": 0.648,
+      "step": 942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6231699029833191,
+      "learning_rate": 9.97631072223796e-06,
+      "loss": 0.5565,
+      "step": 943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8113065507507176,
+      "learning_rate": 9.976210317368436e-06,
+      "loss": 0.6295,
+      "step": 944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6421119369093602,
+      "learning_rate": 9.976109700677537e-06,
+      "loss": 0.5357,
+      "step": 945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6972200604773494,
+      "learning_rate": 9.976008872169552e-06,
+      "loss": 0.6487,
+      "step": 946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6898354852896688,
+      "learning_rate": 9.975907831848768e-06,
+      "loss": 0.5815,
+      "step": 947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.375720273664557,
+      "learning_rate": 9.97580657971949e-06,
+      "loss": 0.6092,
+      "step": 948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8500198409185694,
+      "learning_rate": 9.975705115786025e-06,
+      "loss": 0.6139,
+      "step": 949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8683120723328192,
+      "learning_rate": 9.975603440052694e-06,
+      "loss": 0.6425,
+      "step": 950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.593346455614026,
+      "learning_rate": 9.975501552523822e-06,
+      "loss": 0.5493,
+      "step": 951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8936501859503566,
+      "learning_rate": 9.975399453203752e-06,
+      "loss": 0.6537,
+      "step": 952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9860019505445827,
+      "learning_rate": 9.975297142096825e-06,
+      "loss": 0.7264,
+      "step": 953
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8669851463587487,
+      "learning_rate": 9.975194619207398e-06,
+      "loss": 0.5254,
+      "step": 954
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6487483507891224,
+      "learning_rate": 9.975091884539833e-06,
+      "loss": 0.5469,
+      "step": 955
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7343751264180686,
+      "learning_rate": 9.974988938098505e-06,
+      "loss": 0.6526,
+      "step": 956
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.798451745274211,
+      "learning_rate": 9.974885779887796e-06,
+      "loss": 0.6238,
+      "step": 957
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7944966330683679,
+      "learning_rate": 9.974782409912098e-06,
+      "loss": 0.6414,
+      "step": 958
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7326192905022801,
+      "learning_rate": 9.974678828175808e-06,
+      "loss": 0.5782,
+      "step": 959
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.851409031464957,
+      "learning_rate": 9.974575034683339e-06,
+      "loss": 0.6633,
+      "step": 960
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8329275260160559,
+      "learning_rate": 9.974471029439106e-06,
+      "loss": 0.6178,
+      "step": 961
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8480645643861054,
+      "learning_rate": 9.974366812447538e-06,
+      "loss": 0.6868,
+      "step": 962
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7403624291621149,
+      "learning_rate": 9.974262383713069e-06,
+      "loss": 0.585,
+      "step": 963
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6057429652873727,
+      "learning_rate": 9.974157743240146e-06,
+      "loss": 0.5241,
+      "step": 964
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.723803797580167,
+      "learning_rate": 9.974052891033226e-06,
+      "loss": 0.5509,
+      "step": 965
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7605102234841643,
+      "learning_rate": 9.973947827096766e-06,
+      "loss": 0.6096,
+      "step": 966
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9964860694641393,
+      "learning_rate": 9.973842551435243e-06,
+      "loss": 0.6447,
+      "step": 967
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7125726674264059,
+      "learning_rate": 9.973737064053137e-06,
+      "loss": 0.5714,
+      "step": 968
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6447963992240072,
+      "learning_rate": 9.973631364954937e-06,
+      "loss": 0.5731,
+      "step": 969
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8861228431851509,
+      "learning_rate": 9.973525454145143e-06,
+      "loss": 0.6725,
+      "step": 970
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8599942129085489,
+      "learning_rate": 9.973419331628265e-06,
+      "loss": 0.619,
+      "step": 971
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7870859596350437,
+      "learning_rate": 9.973312997408817e-06,
+      "loss": 0.6642,
+      "step": 972
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8758457278522148,
+      "learning_rate": 9.973206451491329e-06,
+      "loss": 0.6678,
+      "step": 973
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6407520295940252,
+      "learning_rate": 9.973099693880332e-06,
+      "loss": 0.5044,
+      "step": 974
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6075628641057531,
+      "learning_rate": 9.972992724580375e-06,
+      "loss": 0.5654,
+      "step": 975
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.843137238932385,
+      "learning_rate": 9.97288554359601e-06,
+      "loss": 0.692,
+      "step": 976
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8644489293636047,
+      "learning_rate": 9.972778150931797e-06,
+      "loss": 0.6415,
+      "step": 977
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9426617416035463,
+      "learning_rate": 9.972670546592307e-06,
+      "loss": 0.6469,
+      "step": 978
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.64565198646652,
+      "learning_rate": 9.972562730582125e-06,
+      "loss": 0.5428,
+      "step": 979
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6857282054523964,
+      "learning_rate": 9.972454702905837e-06,
+      "loss": 0.5881,
+      "step": 980
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8631417246527289,
+      "learning_rate": 9.97234646356804e-06,
+      "loss": 0.6209,
+      "step": 981
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8870029643336272,
+      "learning_rate": 9.972238012573345e-06,
+      "loss": 0.6494,
+      "step": 982
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6632097527897141,
+      "learning_rate": 9.972129349926368e-06,
+      "loss": 0.5686,
+      "step": 983
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6194036534760328,
+      "learning_rate": 9.972020475631731e-06,
+      "loss": 0.5366,
+      "step": 984
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6992208227502703,
+      "learning_rate": 9.971911389694072e-06,
+      "loss": 0.5421,
+      "step": 985
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.820385052058671,
+      "learning_rate": 9.971802092118032e-06,
+      "loss": 0.6599,
+      "step": 986
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6824016838728466,
+      "learning_rate": 9.971692582908267e-06,
+      "loss": 0.5739,
+      "step": 987
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.898698060195507,
+      "learning_rate": 9.971582862069434e-06,
+      "loss": 0.6914,
+      "step": 988
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7831417966919729,
+      "learning_rate": 9.971472929606206e-06,
+      "loss": 0.637,
+      "step": 989
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8710437767208958,
+      "learning_rate": 9.971362785523261e-06,
+      "loss": 0.6501,
+      "step": 990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9016651239427222,
+      "learning_rate": 9.97125242982529e-06,
+      "loss": 0.6262,
+      "step": 991
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7706646260634279,
+      "learning_rate": 9.971141862516988e-06,
+      "loss": 0.641,
+      "step": 992
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9426023139160854,
+      "learning_rate": 9.971031083603061e-06,
+      "loss": 0.6943,
+      "step": 993
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6896168476557996,
+      "learning_rate": 9.970920093088227e-06,
+      "loss": 0.5623,
+      "step": 994
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9728087762100552,
+      "learning_rate": 9.970808890977211e-06,
+      "loss": 0.7213,
+      "step": 995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9548926293067601,
+      "learning_rate": 9.970697477274744e-06,
+      "loss": 0.6837,
+      "step": 996
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8069373684565228,
+      "learning_rate": 9.970585851985569e-06,
+      "loss": 0.616,
+      "step": 997
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7443443251819689,
+      "learning_rate": 9.970474015114437e-06,
+      "loss": 0.5486,
+      "step": 998
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8556421758965715,
+      "learning_rate": 9.970361966666112e-06,
+      "loss": 0.6649,
+      "step": 999
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6362859396689061,
+      "learning_rate": 9.97024970664536e-06,
+      "loss": 0.5386,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.698702022519884,
+      "learning_rate": 9.97013723505696e-06,
+      "loss": 0.5146,
+      "step": 1001
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.646339997881583,
+      "learning_rate": 9.970024551905701e-06,
+      "loss": 0.5592,
+      "step": 1002
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8423911491600581,
+      "learning_rate": 9.969911657196378e-06,
+      "loss": 0.6253,
+      "step": 1003
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0160592301030402,
+      "learning_rate": 9.969798550933799e-06,
+      "loss": 0.6283,
+      "step": 1004
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.690352142959983,
+      "learning_rate": 9.969685233122774e-06,
+      "loss": 0.5628,
+      "step": 1005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0311787740806562,
+      "learning_rate": 9.969571703768132e-06,
+      "loss": 0.6159,
+      "step": 1006
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7557814251185381,
+      "learning_rate": 9.969457962874702e-06,
+      "loss": 0.5594,
+      "step": 1007
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.785242310763266,
+      "learning_rate": 9.969344010447326e-06,
+      "loss": 0.6654,
+      "step": 1008
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9067553933913962,
+      "learning_rate": 9.969229846490857e-06,
+      "loss": 0.6485,
+      "step": 1009
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6216889007415037,
+      "learning_rate": 9.969115471010152e-06,
+      "loss": 0.5283,
+      "step": 1010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.691634208557814,
+      "learning_rate": 9.96900088401008e-06,
+      "loss": 0.5518,
+      "step": 1011
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7841277537304792,
+      "learning_rate": 9.96888608549552e-06,
+      "loss": 0.6263,
+      "step": 1012
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8179078057475995,
+      "learning_rate": 9.968771075471356e-06,
+      "loss": 0.6032,
+      "step": 1013
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9425520151640354,
+      "learning_rate": 9.968655853942487e-06,
+      "loss": 0.6576,
+      "step": 1014
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8844954600716567,
+      "learning_rate": 9.968540420913815e-06,
+      "loss": 0.628,
+      "step": 1015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8832647227415215,
+      "learning_rate": 9.968424776390254e-06,
+      "loss": 0.6427,
+      "step": 1016
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7438323830426561,
+      "learning_rate": 9.968308920376726e-06,
+      "loss": 0.6119,
+      "step": 1017
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5971458444281293,
+      "learning_rate": 9.968192852878166e-06,
+      "loss": 0.5301,
+      "step": 1018
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8894538345661622,
+      "learning_rate": 9.968076573899513e-06,
+      "loss": 0.6625,
+      "step": 1019
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0120386073613519,
+      "learning_rate": 9.967960083445714e-06,
+      "loss": 0.6678,
+      "step": 1020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9919547610476247,
+      "learning_rate": 9.96784338152173e-06,
+      "loss": 0.7238,
+      "step": 1021
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5978918140517522,
+      "learning_rate": 9.967726468132529e-06,
+      "loss": 0.5507,
+      "step": 1022
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8751885204164667,
+      "learning_rate": 9.967609343283084e-06,
+      "loss": 0.6528,
+      "step": 1023
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6618436429645297,
+      "learning_rate": 9.967492006978386e-06,
+      "loss": 0.5955,
+      "step": 1024
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6316969474791747,
+      "learning_rate": 9.967374459223426e-06,
+      "loss": 0.5325,
+      "step": 1025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.79104689602268,
+      "learning_rate": 9.967256700023212e-06,
+      "loss": 0.5868,
+      "step": 1026
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.011617657288393,
+      "learning_rate": 9.96713872938275e-06,
+      "loss": 0.6631,
+      "step": 1027
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8170905485014265,
+      "learning_rate": 9.967020547307065e-06,
+      "loss": 0.6326,
+      "step": 1028
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.76207282808973,
+      "learning_rate": 9.96690215380119e-06,
+      "loss": 0.6174,
+      "step": 1029
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.72412968908377,
+      "learning_rate": 9.96678354887016e-06,
+      "loss": 0.5365,
+      "step": 1030
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7450700003219325,
+      "learning_rate": 9.966664732519026e-06,
+      "loss": 0.6086,
+      "step": 1031
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9396237903527469,
+      "learning_rate": 9.966545704752845e-06,
+      "loss": 0.6627,
+      "step": 1032
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6340270017350794,
+      "learning_rate": 9.966426465576687e-06,
+      "loss": 0.5418,
+      "step": 1033
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7629589665546378,
+      "learning_rate": 9.966307014995623e-06,
+      "loss": 0.6605,
+      "step": 1034
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6688292179498095,
+      "learning_rate": 9.966187353014739e-06,
+      "loss": 0.5774,
+      "step": 1035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.790214161776146,
+      "learning_rate": 9.96606747963913e-06,
+      "loss": 0.619,
+      "step": 1036
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6345663154255697,
+      "learning_rate": 9.965947394873896e-06,
+      "loss": 0.596,
+      "step": 1037
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7544768452896589,
+      "learning_rate": 9.965827098724152e-06,
+      "loss": 0.6468,
+      "step": 1038
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6907460514008421,
+      "learning_rate": 9.965706591195017e-06,
+      "loss": 0.5965,
+      "step": 1039
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6590134071145285,
+      "learning_rate": 9.96558587229162e-06,
+      "loss": 0.5376,
+      "step": 1040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6689845023180553,
+      "learning_rate": 9.9654649420191e-06,
+      "loss": 0.5715,
+      "step": 1041
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7262232483291121,
+      "learning_rate": 9.965343800382605e-06,
+      "loss": 0.6246,
+      "step": 1042
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.772432713462555,
+      "learning_rate": 9.965222447387291e-06,
+      "loss": 0.5931,
+      "step": 1043
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9080945357905905,
+      "learning_rate": 9.965100883038323e-06,
+      "loss": 0.6755,
+      "step": 1044
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6537616023445731,
+      "learning_rate": 9.964979107340878e-06,
+      "loss": 0.5833,
+      "step": 1045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8274528250485428,
+      "learning_rate": 9.964857120300138e-06,
+      "loss": 0.6862,
+      "step": 1046
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9268645173768608,
+      "learning_rate": 9.964734921921296e-06,
+      "loss": 0.6117,
+      "step": 1047
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.637947616361313,
+      "learning_rate": 9.964612512209553e-06,
+      "loss": 0.56,
+      "step": 1048
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8345885037272234,
+      "learning_rate": 9.964489891170122e-06,
+      "loss": 0.6915,
+      "step": 1049
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.947682923968399,
+      "learning_rate": 9.964367058808217e-06,
+      "loss": 0.557,
+      "step": 1050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6394885257013123,
+      "learning_rate": 9.964244015129071e-06,
+      "loss": 0.5454,
+      "step": 1051
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6039722828318739,
+      "learning_rate": 9.964120760137922e-06,
+      "loss": 0.5202,
+      "step": 1052
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6552510421484344,
+      "learning_rate": 9.963997293840017e-06,
+      "loss": 0.5242,
+      "step": 1053
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9393663003689945,
+      "learning_rate": 9.963873616240607e-06,
+      "loss": 0.6933,
+      "step": 1054
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6995096683237148,
+      "learning_rate": 9.96374972734496e-06,
+      "loss": 0.5895,
+      "step": 1055
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5909357598843684,
+      "learning_rate": 9.96362562715835e-06,
+      "loss": 0.5418,
+      "step": 1056
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8842765522046706,
+      "learning_rate": 9.963501315686057e-06,
+      "loss": 0.6927,
+      "step": 1057
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.769690140762895,
+      "learning_rate": 9.963376792933376e-06,
+      "loss": 0.6504,
+      "step": 1058
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8712214969714248,
+      "learning_rate": 9.963252058905604e-06,
+      "loss": 0.6754,
+      "step": 1059
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7626277381140851,
+      "learning_rate": 9.963127113608054e-06,
+      "loss": 0.58,
+      "step": 1060
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7415564123378365,
+      "learning_rate": 9.963001957046041e-06,
+      "loss": 0.5493,
+      "step": 1061
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8252324047814299,
+      "learning_rate": 9.962876589224894e-06,
+      "loss": 0.6158,
+      "step": 1062
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7740686906153754,
+      "learning_rate": 9.96275101014995e-06,
+      "loss": 0.647,
+      "step": 1063
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6434890865213917,
+      "learning_rate": 9.962625219826554e-06,
+      "loss": 0.5146,
+      "step": 1064
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6356266208073319,
+      "learning_rate": 9.96249921826006e-06,
+      "loss": 0.5051,
+      "step": 1065
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6917109800463024,
+      "learning_rate": 9.962373005455835e-06,
+      "loss": 0.5306,
+      "step": 1066
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6839785692218564,
+      "learning_rate": 9.962246581419246e-06,
+      "loss": 0.599,
+      "step": 1067
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8413408456625441,
+      "learning_rate": 9.962119946155678e-06,
+      "loss": 0.6054,
+      "step": 1068
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8825072950719561,
+      "learning_rate": 9.96199309967052e-06,
+      "loss": 0.6556,
+      "step": 1069
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6568332605989884,
+      "learning_rate": 9.961866041969172e-06,
+      "loss": 0.5804,
+      "step": 1070
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6701904019859587,
+      "learning_rate": 9.961738773057044e-06,
+      "loss": 0.5573,
+      "step": 1071
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8403411171391452,
+      "learning_rate": 9.96161129293955e-06,
+      "loss": 0.5507,
+      "step": 1072
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6320571860344364,
+      "learning_rate": 9.96148360162212e-06,
+      "loss": 0.4739,
+      "step": 1073
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8241160117697117,
+      "learning_rate": 9.961355699110188e-06,
+      "loss": 0.6369,
+      "step": 1074
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6775006836052057,
+      "learning_rate": 9.961227585409194e-06,
+      "loss": 0.5706,
+      "step": 1075
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.782967568295276,
+      "learning_rate": 9.961099260524601e-06,
+      "loss": 0.6293,
+      "step": 1076
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9922788598511965,
+      "learning_rate": 9.960970724461862e-06,
+      "loss": 0.6214,
+      "step": 1077
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6835477547128451,
+      "learning_rate": 9.960841977226455e-06,
+      "loss": 0.5708,
+      "step": 1078
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7085800777580535,
+      "learning_rate": 9.960713018823855e-06,
+      "loss": 0.5775,
+      "step": 1079
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8834974525895641,
+      "learning_rate": 9.960583849259556e-06,
+      "loss": 0.6351,
+      "step": 1080
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7213832710568305,
+      "learning_rate": 9.960454468539053e-06,
+      "loss": 0.5675,
+      "step": 1081
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8879443734615274,
+      "learning_rate": 9.960324876667854e-06,
+      "loss": 0.6596,
+      "step": 1082
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7344950907712255,
+      "learning_rate": 9.960195073651478e-06,
+      "loss": 0.565,
+      "step": 1083
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7007327711779928,
+      "learning_rate": 9.960065059495446e-06,
+      "loss": 0.5203,
+      "step": 1084
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9642709628559569,
+      "learning_rate": 9.959934834205296e-06,
+      "loss": 0.5881,
+      "step": 1085
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7909316044844384,
+      "learning_rate": 9.95980439778657e-06,
+      "loss": 0.5805,
+      "step": 1086
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6601031474911905,
+      "learning_rate": 9.95967375024482e-06,
+      "loss": 0.5908,
+      "step": 1087
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7576039583517902,
+      "learning_rate": 9.959542891585606e-06,
+      "loss": 0.5639,
+      "step": 1088
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9212523958223972,
+      "learning_rate": 9.9594118218145e-06,
+      "loss": 0.6982,
+      "step": 1089
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5924096392485172,
+      "learning_rate": 9.959280540937082e-06,
+      "loss": 0.5012,
+      "step": 1090
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6641663266900463,
+      "learning_rate": 9.959149048958938e-06,
+      "loss": 0.5456,
+      "step": 1091
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9091526004034339,
+      "learning_rate": 9.959017345885666e-06,
+      "loss": 0.7084,
+      "step": 1092
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1011575016214283,
+      "learning_rate": 9.958885431722874e-06,
+      "loss": 0.6814,
+      "step": 1093
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9750371123627987,
+      "learning_rate": 9.958753306476172e-06,
+      "loss": 0.6081,
+      "step": 1094
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6333414917853868,
+      "learning_rate": 9.95862097015119e-06,
+      "loss": 0.5111,
+      "step": 1095
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6731356071365201,
+      "learning_rate": 9.95848842275356e-06,
+      "loss": 0.5338,
+      "step": 1096
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8756458617676428,
+      "learning_rate": 9.95835566428892e-06,
+      "loss": 0.5627,
+      "step": 1097
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7144933715655397,
+      "learning_rate": 9.958222694762926e-06,
+      "loss": 0.6085,
+      "step": 1098
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8363796515389963,
+      "learning_rate": 9.958089514181236e-06,
+      "loss": 0.586,
+      "step": 1099
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.027379405617287,
+      "learning_rate": 9.95795612254952e-06,
+      "loss": 0.6114,
+      "step": 1100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0292639883298103,
+      "learning_rate": 9.957822519873453e-06,
+      "loss": 0.6284,
+      "step": 1101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0574842183994442,
+      "learning_rate": 9.957688706158725e-06,
+      "loss": 0.5952,
+      "step": 1102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8483335753387509,
+      "learning_rate": 9.957554681411032e-06,
+      "loss": 0.6091,
+      "step": 1103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7428902607788493,
+      "learning_rate": 9.957420445636077e-06,
+      "loss": 0.6634,
+      "step": 1104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7026104738413125,
+      "learning_rate": 9.957285998839577e-06,
+      "loss": 0.5774,
+      "step": 1105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7245714299947138,
+      "learning_rate": 9.957151341027251e-06,
+      "loss": 0.6126,
+      "step": 1106
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5621554383607955,
+      "learning_rate": 9.957016472204834e-06,
+      "loss": 0.5149,
+      "step": 1107
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.848785091007283,
+      "learning_rate": 9.956881392378068e-06,
+      "loss": 0.6135,
+      "step": 1108
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6024732942844954,
+      "learning_rate": 9.9567461015527e-06,
+      "loss": 0.5038,
+      "step": 1109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7577805467470974,
+      "learning_rate": 9.956610599734488e-06,
+      "loss": 0.5623,
+      "step": 1110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6529702214325929,
+      "learning_rate": 9.956474886929205e-06,
+      "loss": 0.5759,
+      "step": 1111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8385138547411434,
+      "learning_rate": 9.956338963142622e-06,
+      "loss": 0.6259,
+      "step": 1112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6348340774981751,
+      "learning_rate": 9.95620282838053e-06,
+      "loss": 0.5491,
+      "step": 1113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8432930975231842,
+      "learning_rate": 9.95606648264872e-06,
+      "loss": 0.6528,
+      "step": 1114
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.646499926993693,
+      "learning_rate": 9.955929925952996e-06,
+      "loss": 0.5908,
+      "step": 1115
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.829058950554376,
+      "learning_rate": 9.955793158299173e-06,
+      "loss": 0.6599,
+      "step": 1116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.613515823068835,
+      "learning_rate": 9.955656179693073e-06,
+      "loss": 0.4962,
+      "step": 1117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.090777028471484,
+      "learning_rate": 9.955518990140525e-06,
+      "loss": 0.6454,
+      "step": 1118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5946182803339479,
+      "learning_rate": 9.955381589647367e-06,
+      "loss": 0.5602,
+      "step": 1119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8286584023219346,
+      "learning_rate": 9.955243978219452e-06,
+      "loss": 0.6051,
+      "step": 1120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6467996168545798,
+      "learning_rate": 9.955106155862635e-06,
+      "loss": 0.5473,
+      "step": 1121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5940894978116646,
+      "learning_rate": 9.954968122582784e-06,
+      "loss": 0.5338,
+      "step": 1122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8590824308861795,
+      "learning_rate": 9.954829878385773e-06,
+      "loss": 0.6326,
+      "step": 1123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6244978334185487,
+      "learning_rate": 9.954691423277487e-06,
+      "loss": 0.5246,
+      "step": 1124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5909229083578984,
+      "learning_rate": 9.95455275726382e-06,
+      "loss": 0.5258,
+      "step": 1125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7924079847665105,
+      "learning_rate": 9.954413880350674e-06,
+      "loss": 0.6057,
+      "step": 1126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.008750003158618,
+      "learning_rate": 9.954274792543963e-06,
+      "loss": 0.6279,
+      "step": 1127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9275739014823893,
+      "learning_rate": 9.954135493849605e-06,
+      "loss": 0.6504,
+      "step": 1128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6588408665126066,
+      "learning_rate": 9.95399598427353e-06,
+      "loss": 0.5735,
+      "step": 1129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7338741263032064,
+      "learning_rate": 9.953856263821677e-06,
+      "loss": 0.536,
+      "step": 1130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6913121183957724,
+      "learning_rate": 9.953716332499991e-06,
+      "loss": 0.6244,
+      "step": 1131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9019267806465207,
+      "learning_rate": 9.953576190314434e-06,
+      "loss": 0.6787,
+      "step": 1132
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6933534279255132,
+      "learning_rate": 9.953435837270966e-06,
+      "loss": 0.5362,
+      "step": 1133
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8972807751072782,
+      "learning_rate": 9.953295273375564e-06,
+      "loss": 0.6749,
+      "step": 1134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6176012154418374,
+      "learning_rate": 9.95315449863421e-06,
+      "loss": 0.5268,
+      "step": 1135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9166086004029897,
+      "learning_rate": 9.953013513052898e-06,
+      "loss": 0.5797,
+      "step": 1136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7726218336683778,
+      "learning_rate": 9.952872316637627e-06,
+      "loss": 0.593,
+      "step": 1137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6059201439858781,
+      "learning_rate": 9.952730909394409e-06,
+      "loss": 0.5262,
+      "step": 1138
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8709987021880209,
+      "learning_rate": 9.952589291329262e-06,
+      "loss": 0.6675,
+      "step": 1139
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9166989103261511,
+      "learning_rate": 9.952447462448219e-06,
+      "loss": 0.6499,
+      "step": 1140
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7545998650211926,
+      "learning_rate": 9.952305422757308e-06,
+      "loss": 0.6138,
+      "step": 1141
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2410058081739495,
+      "learning_rate": 9.952163172262583e-06,
+      "loss": 0.6333,
+      "step": 1142
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7992091222627921,
+      "learning_rate": 9.952020710970098e-06,
+      "loss": 0.5316,
+      "step": 1143
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7289102533460955,
+      "learning_rate": 9.951878038885914e-06,
+      "loss": 0.537,
+      "step": 1144
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6614206102741648,
+      "learning_rate": 9.951735156016105e-06,
+      "loss": 0.5623,
+      "step": 1145
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8177142563654926,
+      "learning_rate": 9.951592062366754e-06,
+      "loss": 0.6382,
+      "step": 1146
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7097636836053219,
+      "learning_rate": 9.951448757943954e-06,
+      "loss": 0.5972,
+      "step": 1147
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9254677182285955,
+      "learning_rate": 9.951305242753801e-06,
+      "loss": 0.571,
+      "step": 1148
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6988956392409752,
+      "learning_rate": 9.951161516802408e-06,
+      "loss": 0.5549,
+      "step": 1149
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8134103090839735,
+      "learning_rate": 9.951017580095889e-06,
+      "loss": 0.5931,
+      "step": 1150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6059200480854988,
+      "learning_rate": 9.950873432640373e-06,
+      "loss": 0.5429,
+      "step": 1151
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9848688757575335,
+      "learning_rate": 9.950729074441998e-06,
+      "loss": 0.6748,
+      "step": 1152
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5763030200172503,
+      "learning_rate": 9.950584505506904e-06,
+      "loss": 0.495,
+      "step": 1153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7517038459961862,
+      "learning_rate": 9.950439725841247e-06,
+      "loss": 0.5802,
+      "step": 1154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0629583474210524,
+      "learning_rate": 9.950294735451192e-06,
+      "loss": 0.6841,
+      "step": 1155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7663206359698448,
+      "learning_rate": 9.950149534342907e-06,
+      "loss": 0.5319,
+      "step": 1156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7795779642312519,
+      "learning_rate": 9.950004122522578e-06,
+      "loss": 0.6035,
+      "step": 1157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.178206918510157,
+      "learning_rate": 9.949858499996389e-06,
+      "loss": 0.6628,
+      "step": 1158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7391505252721015,
+      "learning_rate": 9.949712666770541e-06,
+      "loss": 0.5569,
+      "step": 1159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7583907860065018,
+      "learning_rate": 9.949566622851243e-06,
+      "loss": 0.5678,
+      "step": 1160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.646869541150209,
+      "learning_rate": 9.94942036824471e-06,
+      "loss": 0.5737,
+      "step": 1161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.794371583641433,
+      "learning_rate": 9.949273902957169e-06,
+      "loss": 0.6565,
+      "step": 1162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9050968642999875,
+      "learning_rate": 9.94912722699485e-06,
+      "loss": 0.5778,
+      "step": 1163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6189268621262112,
+      "learning_rate": 9.948980340364002e-06,
+      "loss": 0.5335,
+      "step": 1164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6192781388637013,
+      "learning_rate": 9.948833243070877e-06,
+      "loss": 0.5795,
+      "step": 1165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6997940420230017,
+      "learning_rate": 9.948685935121735e-06,
+      "loss": 0.5817,
+      "step": 1166
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.666587408485315,
+      "learning_rate": 9.948538416522846e-06,
+      "loss": 0.5358,
+      "step": 1167
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.922220574123095,
+      "learning_rate": 9.948390687280489e-06,
+      "loss": 0.6504,
+      "step": 1168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9291835324778818,
+      "learning_rate": 9.948242747400953e-06,
+      "loss": 0.6552,
+      "step": 1169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6609464921861429,
+      "learning_rate": 9.948094596890536e-06,
+      "loss": 0.5258,
+      "step": 1170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7432722982339321,
+      "learning_rate": 9.947946235755545e-06,
+      "loss": 0.518,
+      "step": 1171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9232759463515889,
+      "learning_rate": 9.947797664002294e-06,
+      "loss": 0.6911,
+      "step": 1172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9677536032809348,
+      "learning_rate": 9.947648881637107e-06,
+      "loss": 0.6149,
+      "step": 1173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0779556679380788,
+      "learning_rate": 9.947499888666317e-06,
+      "loss": 0.6383,
+      "step": 1174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6818092950991045,
+      "learning_rate": 9.947350685096266e-06,
+      "loss": 0.5611,
+      "step": 1175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6562011194061566,
+      "learning_rate": 9.947201270933307e-06,
+      "loss": 0.558,
+      "step": 1176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7506221775353435,
+      "learning_rate": 9.947051646183798e-06,
+      "loss": 0.6525,
+      "step": 1177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9732030857141276,
+      "learning_rate": 9.946901810854109e-06,
+      "loss": 0.6358,
+      "step": 1178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6719722587021625,
+      "learning_rate": 9.94675176495062e-06,
+      "loss": 0.5724,
+      "step": 1179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7277819505454043,
+      "learning_rate": 9.946601508479714e-06,
+      "loss": 0.5659,
+      "step": 1180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6024131898213151,
+      "learning_rate": 9.946451041447788e-06,
+      "loss": 0.5407,
+      "step": 1181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6194888085588076,
+      "learning_rate": 9.94630036386125e-06,
+      "loss": 0.531,
+      "step": 1182
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.819612771311822,
+      "learning_rate": 9.946149475726509e-06,
+      "loss": 0.6299,
+      "step": 1183
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5988162409384846,
+      "learning_rate": 9.945998377049992e-06,
+      "loss": 0.5163,
+      "step": 1184
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.621047731739097,
+      "learning_rate": 9.945847067838131e-06,
+      "loss": 0.6432,
+      "step": 1185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6933969601897686,
+      "learning_rate": 9.945695548097363e-06,
+      "loss": 0.55,
+      "step": 1186
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6215097677045021,
+      "learning_rate": 9.94554381783414e-06,
+      "loss": 0.627,
+      "step": 1187
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9398154868265022,
+      "learning_rate": 9.945391877054919e-06,
+      "loss": 0.7062,
+      "step": 1188
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9323473338926082,
+      "learning_rate": 9.945239725766172e-06,
+      "loss": 0.6333,
+      "step": 1189
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9423907671305878,
+      "learning_rate": 9.94508736397437e-06,
+      "loss": 0.6752,
+      "step": 1190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9615586928806058,
+      "learning_rate": 9.944934791686003e-06,
+      "loss": 0.6944,
+      "step": 1191
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6415203280411638,
+      "learning_rate": 9.944782008907564e-06,
+      "loss": 0.5337,
+      "step": 1192
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6503610631443516,
+      "learning_rate": 9.944629015645553e-06,
+      "loss": 0.517,
+      "step": 1193
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.189574393483398,
+      "learning_rate": 9.944475811906487e-06,
+      "loss": 0.6663,
+      "step": 1194
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8127292741305696,
+      "learning_rate": 9.944322397696888e-06,
+      "loss": 0.6214,
+      "step": 1195
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9352628663178603,
+      "learning_rate": 9.944168773023282e-06,
+      "loss": 0.6445,
+      "step": 1196
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8159532544926157,
+      "learning_rate": 9.944014937892211e-06,
+      "loss": 0.6347,
+      "step": 1197
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7529396903292472,
+      "learning_rate": 9.943860892310225e-06,
+      "loss": 0.5708,
+      "step": 1198
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7332699396889881,
+      "learning_rate": 9.943706636283876e-06,
+      "loss": 0.617,
+      "step": 1199
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7065292946089605,
+      "learning_rate": 9.943552169819734e-06,
+      "loss": 0.5599,
+      "step": 1200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9523616116392661,
+      "learning_rate": 9.943397492924377e-06,
+      "loss": 0.6991,
+      "step": 1201
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6718897182036292,
+      "learning_rate": 9.943242605604381e-06,
+      "loss": 0.5334,
+      "step": 1202
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6255929956273176,
+      "learning_rate": 9.943087507866345e-06,
+      "loss": 0.5123,
+      "step": 1203
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5863924057030013,
+      "learning_rate": 9.94293219971687e-06,
+      "loss": 0.5259,
+      "step": 1204
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6129535666687689,
+      "learning_rate": 9.942776681162566e-06,
+      "loss": 0.5341,
+      "step": 1205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6738433228662217,
+      "learning_rate": 9.942620952210057e-06,
+      "loss": 0.5877,
+      "step": 1206
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6691527608264151,
+      "learning_rate": 9.942465012865964e-06,
+      "loss": 0.6106,
+      "step": 1207
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8947966968433815,
+      "learning_rate": 9.94230886313693e-06,
+      "loss": 0.6766,
+      "step": 1208
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5937059305694171,
+      "learning_rate": 9.942152503029603e-06,
+      "loss": 0.5424,
+      "step": 1209
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6666550628105659,
+      "learning_rate": 9.941995932550636e-06,
+      "loss": 0.6206,
+      "step": 1210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6855403367079185,
+      "learning_rate": 9.941839151706694e-06,
+      "loss": 0.5557,
+      "step": 1211
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6996973493623373,
+      "learning_rate": 9.941682160504452e-06,
+      "loss": 0.638,
+      "step": 1212
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8022251873817732,
+      "learning_rate": 9.941524958950591e-06,
+      "loss": 0.6535,
+      "step": 1213
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8575600904950452,
+      "learning_rate": 9.941367547051803e-06,
+      "loss": 0.5889,
+      "step": 1214
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7458489001018135,
+      "learning_rate": 9.94120992481479e-06,
+      "loss": 0.6282,
+      "step": 1215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7011274758384984,
+      "learning_rate": 9.94105209224626e-06,
+      "loss": 0.6035,
+      "step": 1216
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.778022051851357,
+      "learning_rate": 9.940894049352932e-06,
+      "loss": 0.6024,
+      "step": 1217
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8229848811812075,
+      "learning_rate": 9.940735796141533e-06,
+      "loss": 0.7138,
+      "step": 1218
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6693289356258347,
+      "learning_rate": 9.940577332618798e-06,
+      "loss": 0.6005,
+      "step": 1219
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8768519648063721,
+      "learning_rate": 9.940418658791475e-06,
+      "loss": 0.6364,
+      "step": 1220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6210843299179503,
+      "learning_rate": 9.940259774666316e-06,
+      "loss": 0.5151,
+      "step": 1221
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5928265232338573,
+      "learning_rate": 9.940100680250086e-06,
+      "loss": 0.5706,
+      "step": 1222
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7666455691407895,
+      "learning_rate": 9.939941375549559e-06,
+      "loss": 0.611,
+      "step": 1223
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8285991798299784,
+      "learning_rate": 9.93978186057151e-06,
+      "loss": 0.657,
+      "step": 1224
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5513770793701552,
+      "learning_rate": 9.939622135322733e-06,
+      "loss": 0.5328,
+      "step": 1225
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6724169105488185,
+      "learning_rate": 9.939462199810027e-06,
+      "loss": 0.5896,
+      "step": 1226
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6217065032154994,
+      "learning_rate": 9.9393020540402e-06,
+      "loss": 0.535,
+      "step": 1227
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.714592607405014,
+      "learning_rate": 9.939141698020067e-06,
+      "loss": 0.5438,
+      "step": 1228
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6270973608896668,
+      "learning_rate": 9.938981131756455e-06,
+      "loss": 0.5692,
+      "step": 1229
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7516987047795822,
+      "learning_rate": 9.938820355256201e-06,
+      "loss": 0.6568,
+      "step": 1230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6853617324725388,
+      "learning_rate": 9.938659368526146e-06,
+      "loss": 0.5389,
+      "step": 1231
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.739056326079907,
+      "learning_rate": 9.938498171573142e-06,
+      "loss": 0.5849,
+      "step": 1232
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8286906199401044,
+      "learning_rate": 9.938336764404053e-06,
+      "loss": 0.6549,
+      "step": 1233
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6871217548760605,
+      "learning_rate": 9.938175147025749e-06,
+      "loss": 0.5441,
+      "step": 1234
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.635761385045518,
+      "learning_rate": 9.938013319445107e-06,
+      "loss": 0.5623,
+      "step": 1235
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7732509643778649,
+      "learning_rate": 9.93785128166902e-06,
+      "loss": 0.6476,
+      "step": 1236
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9388942019799706,
+      "learning_rate": 9.937689033704383e-06,
+      "loss": 0.6313,
+      "step": 1237
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6506073910645148,
+      "learning_rate": 9.937526575558102e-06,
+      "loss": 0.5507,
+      "step": 1238
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7109459417537408,
+      "learning_rate": 9.937363907237093e-06,
+      "loss": 0.5455,
+      "step": 1239
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7131945189513951,
+      "learning_rate": 9.937201028748278e-06,
+      "loss": 0.5972,
+      "step": 1240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7933921050202154,
+      "learning_rate": 9.937037940098595e-06,
+      "loss": 0.6384,
+      "step": 1241
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7117828298931219,
+      "learning_rate": 9.936874641294982e-06,
+      "loss": 0.5741,
+      "step": 1242
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6858607363442373,
+      "learning_rate": 9.936711132344393e-06,
+      "loss": 0.5684,
+      "step": 1243
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.079827994215085,
+      "learning_rate": 9.936547413253783e-06,
+      "loss": 0.6917,
+      "step": 1244
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8693329698611052,
+      "learning_rate": 9.936383484030127e-06,
+      "loss": 0.6281,
+      "step": 1245
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9237381724365175,
+      "learning_rate": 9.9362193446804e-06,
+      "loss": 0.6816,
+      "step": 1246
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5947942369877502,
+      "learning_rate": 9.93605499521159e-06,
+      "loss": 0.5281,
+      "step": 1247
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6876185476635489,
+      "learning_rate": 9.935890435630693e-06,
+      "loss": 0.5819,
+      "step": 1248
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7852566555702422,
+      "learning_rate": 9.935725665944712e-06,
+      "loss": 0.5835,
+      "step": 1249
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8860897748537813,
+      "learning_rate": 9.935560686160661e-06,
+      "loss": 0.6229,
+      "step": 1250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8380467170046101,
+      "learning_rate": 9.935395496285565e-06,
+      "loss": 0.5754,
+      "step": 1251
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8174419125773005,
+      "learning_rate": 9.935230096326452e-06,
+      "loss": 0.5831,
+      "step": 1252
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6096272258583538,
+      "learning_rate": 9.935064486290366e-06,
+      "loss": 0.5627,
+      "step": 1253
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8144193664929303,
+      "learning_rate": 9.934898666184354e-06,
+      "loss": 0.6342,
+      "step": 1254
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6618055773716802,
+      "learning_rate": 9.934732636015475e-06,
+      "loss": 0.558,
+      "step": 1255
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5670649211847698,
+      "learning_rate": 9.934566395790798e-06,
+      "loss": 0.5249,
+      "step": 1256
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0116605711057045,
+      "learning_rate": 9.934399945517398e-06,
+      "loss": 0.7115,
+      "step": 1257
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9606958544754229,
+      "learning_rate": 9.934233285202362e-06,
+      "loss": 0.6245,
+      "step": 1258
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6005659704302857,
+      "learning_rate": 9.93406641485278e-06,
+      "loss": 0.5508,
+      "step": 1259
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.603908805834337,
+      "learning_rate": 9.93389933447576e-06,
+      "loss": 0.5268,
+      "step": 1260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6912683190406798,
+      "learning_rate": 9.93373204407841e-06,
+      "loss": 0.5603,
+      "step": 1261
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5015099086663268,
+      "learning_rate": 9.933564543667854e-06,
+      "loss": 0.6758,
+      "step": 1262
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.61498877277781,
+      "learning_rate": 9.933396833251221e-06,
+      "loss": 0.5979,
+      "step": 1263
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7587062308638808,
+      "learning_rate": 9.933228912835649e-06,
+      "loss": 0.535,
+      "step": 1264
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6372135461611754,
+      "learning_rate": 9.933060782428286e-06,
+      "loss": 0.5916,
+      "step": 1265
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8828201367862585,
+      "learning_rate": 9.932892442036289e-06,
+      "loss": 0.5955,
+      "step": 1266
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8675700039740134,
+      "learning_rate": 9.932723891666825e-06,
+      "loss": 0.6343,
+      "step": 1267
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.000392034730579,
+      "learning_rate": 9.932555131327069e-06,
+      "loss": 0.5913,
+      "step": 1268
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6441274395889746,
+      "learning_rate": 9.9323861610242e-06,
+      "loss": 0.5294,
+      "step": 1269
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5634681746022322,
+      "learning_rate": 9.932216980765416e-06,
+      "loss": 0.5094,
+      "step": 1270
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8087186209107541,
+      "learning_rate": 9.932047590557916e-06,
+      "loss": 0.6285,
+      "step": 1271
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5830632100716865,
+      "learning_rate": 9.93187799040891e-06,
+      "loss": 0.5117,
+      "step": 1272
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5993417802328452,
+      "learning_rate": 9.931708180325619e-06,
+      "loss": 0.5062,
+      "step": 1273
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.669294113898828,
+      "learning_rate": 9.931538160315268e-06,
+      "loss": 0.5355,
+      "step": 1274
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.60490174747613,
+      "learning_rate": 9.931367930385098e-06,
+      "loss": 0.5085,
+      "step": 1275
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6262615291290174,
+      "learning_rate": 9.931197490542354e-06,
+      "loss": 0.5536,
+      "step": 1276
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.91396197589831,
+      "learning_rate": 9.931026840794292e-06,
+      "loss": 0.6301,
+      "step": 1277
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0022672535619206,
+      "learning_rate": 9.930855981148172e-06,
+      "loss": 0.6549,
+      "step": 1278
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9263003202148294,
+      "learning_rate": 9.93068491161127e-06,
+      "loss": 0.6036,
+      "step": 1279
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7871313319089243,
+      "learning_rate": 9.930513632190868e-06,
+      "loss": 0.643,
+      "step": 1280
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7385298884516197,
+      "learning_rate": 9.930342142894259e-06,
+      "loss": 0.544,
+      "step": 1281
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9343110234348226,
+      "learning_rate": 9.930170443728736e-06,
+      "loss": 0.6865,
+      "step": 1282
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.778588128833451,
+      "learning_rate": 9.929998534701612e-06,
+      "loss": 0.6069,
+      "step": 1283
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0524259183173499,
+      "learning_rate": 9.929826415820207e-06,
+      "loss": 0.6236,
+      "step": 1284
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8147387322477797,
+      "learning_rate": 9.929654087091845e-06,
+      "loss": 0.5876,
+      "step": 1285
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8626286053173774,
+      "learning_rate": 9.92948154852386e-06,
+      "loss": 0.6699,
+      "step": 1286
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6848567231390712,
+      "learning_rate": 9.929308800123597e-06,
+      "loss": 0.5203,
+      "step": 1287
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8106992091885179,
+      "learning_rate": 9.929135841898412e-06,
+      "loss": 0.6333,
+      "step": 1288
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6355507244372836,
+      "learning_rate": 9.928962673855664e-06,
+      "loss": 0.5674,
+      "step": 1289
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9575810776483066,
+      "learning_rate": 9.928789296002726e-06,
+      "loss": 0.5601,
+      "step": 1290
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6614320788002571,
+      "learning_rate": 9.928615708346978e-06,
+      "loss": 0.5323,
+      "step": 1291
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.78603144821625,
+      "learning_rate": 9.92844191089581e-06,
+      "loss": 0.63,
+      "step": 1292
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.854145474325908,
+      "learning_rate": 9.92826790365662e-06,
+      "loss": 0.5931,
+      "step": 1293
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9479017928633185,
+      "learning_rate": 9.928093686636811e-06,
+      "loss": 0.6366,
+      "step": 1294
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6556789536313667,
+      "learning_rate": 9.927919259843801e-06,
+      "loss": 0.6144,
+      "step": 1295
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7178000902835753,
+      "learning_rate": 9.927744623285017e-06,
+      "loss": 0.6428,
+      "step": 1296
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9778528421679374,
+      "learning_rate": 9.927569776967891e-06,
+      "loss": 0.6789,
+      "step": 1297
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.699409897187686,
+      "learning_rate": 9.927394720899866e-06,
+      "loss": 0.5989,
+      "step": 1298
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7475897085800394,
+      "learning_rate": 9.927219455088394e-06,
+      "loss": 0.5351,
+      "step": 1299
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9380606806438925,
+      "learning_rate": 9.927043979540934e-06,
+      "loss": 0.6808,
+      "step": 1300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7623954190929098,
+      "learning_rate": 9.926868294264957e-06,
+      "loss": 0.6033,
+      "step": 1301
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8451633671559466,
+      "learning_rate": 9.92669239926794e-06,
+      "loss": 0.6047,
+      "step": 1302
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6344557359348485,
+      "learning_rate": 9.926516294557374e-06,
+      "loss": 0.4629,
+      "step": 1303
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.983505389962346,
+      "learning_rate": 9.92633998014075e-06,
+      "loss": 0.6401,
+      "step": 1304
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8396107779558755,
+      "learning_rate": 9.926163456025573e-06,
+      "loss": 0.5932,
+      "step": 1305
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6464848349421851,
+      "learning_rate": 9.925986722219362e-06,
+      "loss": 0.5482,
+      "step": 1306
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6152424291581389,
+      "learning_rate": 9.925809778729639e-06,
+      "loss": 0.5177,
+      "step": 1307
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8963702088469717,
+      "learning_rate": 9.925632625563931e-06,
+      "loss": 0.6215,
+      "step": 1308
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8236919014857446,
+      "learning_rate": 9.925455262729785e-06,
+      "loss": 0.6344,
+      "step": 1309
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8067235544500863,
+      "learning_rate": 9.925277690234745e-06,
+      "loss": 0.6568,
+      "step": 1310
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8053307649416414,
+      "learning_rate": 9.925099908086375e-06,
+      "loss": 0.6914,
+      "step": 1311
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8185646572828913,
+      "learning_rate": 9.924921916292239e-06,
+      "loss": 0.6134,
+      "step": 1312
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8488052935040817,
+      "learning_rate": 9.924743714859915e-06,
+      "loss": 0.6114,
+      "step": 1313
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.786538209385159,
+      "learning_rate": 9.924565303796988e-06,
+      "loss": 0.6325,
+      "step": 1314
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8421790302765904,
+      "learning_rate": 9.924386683111055e-06,
+      "loss": 0.6299,
+      "step": 1315
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8593619423668151,
+      "learning_rate": 9.924207852809715e-06,
+      "loss": 0.5854,
+      "step": 1316
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7794165596221101,
+      "learning_rate": 9.924028812900582e-06,
+      "loss": 0.6472,
+      "step": 1317
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6200289391608015,
+      "learning_rate": 9.923849563391278e-06,
+      "loss": 0.5082,
+      "step": 1318
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6503282906773031,
+      "learning_rate": 9.923670104289433e-06,
+      "loss": 0.5554,
+      "step": 1319
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9334848680283766,
+      "learning_rate": 9.923490435602685e-06,
+      "loss": 0.6438,
+      "step": 1320
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.853719193099155,
+      "learning_rate": 9.923310557338681e-06,
+      "loss": 0.5958,
+      "step": 1321
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6323711898369337,
+      "learning_rate": 9.923130469505081e-06,
+      "loss": 0.524,
+      "step": 1322
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6266910281118175,
+      "learning_rate": 9.922950172109549e-06,
+      "loss": 0.5212,
+      "step": 1323
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6949655797370283,
+      "learning_rate": 9.922769665159759e-06,
+      "loss": 0.5901,
+      "step": 1324
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8504757543505016,
+      "learning_rate": 9.922588948663395e-06,
+      "loss": 0.6324,
+      "step": 1325
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.198837872795737,
+      "learning_rate": 9.92240802262815e-06,
+      "loss": 0.5773,
+      "step": 1326
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7389168502790365,
+      "learning_rate": 9.922226887061726e-06,
+      "loss": 0.6019,
+      "step": 1327
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9543392867703335,
+      "learning_rate": 9.922045541971831e-06,
+      "loss": 0.6718,
+      "step": 1328
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6769447419107208,
+      "learning_rate": 9.921863987366187e-06,
+      "loss": 0.5451,
+      "step": 1329
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7995115583222592,
+      "learning_rate": 9.921682223252522e-06,
+      "loss": 0.6461,
+      "step": 1330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5807904876349749,
+      "learning_rate": 9.921500249638572e-06,
+      "loss": 0.5647,
+      "step": 1331
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7328444051173592,
+      "learning_rate": 9.921318066532082e-06,
+      "loss": 0.5567,
+      "step": 1332
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6423699956731261,
+      "learning_rate": 9.921135673940809e-06,
+      "loss": 0.532,
+      "step": 1333
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0732783771120218,
+      "learning_rate": 9.920953071872514e-06,
+      "loss": 0.6393,
+      "step": 1334
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6072194907461469,
+      "learning_rate": 9.920770260334974e-06,
+      "loss": 0.5242,
+      "step": 1335
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8678831232928724,
+      "learning_rate": 9.920587239335968e-06,
+      "loss": 0.6428,
+      "step": 1336
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7322646257079967,
+      "learning_rate": 9.920404008883287e-06,
+      "loss": 0.582,
+      "step": 1337
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6987065357589037,
+      "learning_rate": 9.920220568984731e-06,
+      "loss": 0.5263,
+      "step": 1338
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.894246119987906,
+      "learning_rate": 9.920036919648108e-06,
+      "loss": 0.6545,
+      "step": 1339
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7271150938938604,
+      "learning_rate": 9.919853060881238e-06,
+      "loss": 0.6291,
+      "step": 1340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0716868696461448,
+      "learning_rate": 9.919668992691941e-06,
+      "loss": 0.6191,
+      "step": 1341
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.07794875912514,
+      "learning_rate": 9.919484715088057e-06,
+      "loss": 0.6533,
+      "step": 1342
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7029692021347415,
+      "learning_rate": 9.91930022807743e-06,
+      "loss": 0.5706,
+      "step": 1343
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5813230784025685,
+      "learning_rate": 9.919115531667911e-06,
+      "loss": 0.5593,
+      "step": 1344
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.605113227667247,
+      "learning_rate": 9.918930625867364e-06,
+      "loss": 0.5373,
+      "step": 1345
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7225811940174828,
+      "learning_rate": 9.918745510683659e-06,
+      "loss": 0.6309,
+      "step": 1346
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6053881124764692,
+      "learning_rate": 9.918560186124674e-06,
+      "loss": 0.5351,
+      "step": 1347
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6818319116848462,
+      "learning_rate": 9.918374652198302e-06,
+      "loss": 0.5734,
+      "step": 1348
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8958174067084802,
+      "learning_rate": 9.918188908912436e-06,
+      "loss": 0.6048,
+      "step": 1349
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8187811101690751,
+      "learning_rate": 9.918002956274986e-06,
+      "loss": 0.6245,
+      "step": 1350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7363704720909162,
+      "learning_rate": 9.917816794293864e-06,
+      "loss": 0.5962,
+      "step": 1351
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7539830444599103,
+      "learning_rate": 9.917630422976997e-06,
+      "loss": 0.632,
+      "step": 1352
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.800461503275176,
+      "learning_rate": 9.917443842332318e-06,
+      "loss": 0.6383,
+      "step": 1353
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5859864768068026,
+      "learning_rate": 9.917257052367768e-06,
+      "loss": 0.5852,
+      "step": 1354
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7748718361879976,
+      "learning_rate": 9.917070053091298e-06,
+      "loss": 0.5468,
+      "step": 1355
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7480038777585786,
+      "learning_rate": 9.916882844510868e-06,
+      "loss": 0.6377,
+      "step": 1356
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5950072074474939,
+      "learning_rate": 9.91669542663445e-06,
+      "loss": 0.5946,
+      "step": 1357
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7547818256118963,
+      "learning_rate": 9.916507799470016e-06,
+      "loss": 0.5945,
+      "step": 1358
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6190436223612008,
+      "learning_rate": 9.916319963025558e-06,
+      "loss": 0.588,
+      "step": 1359
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6289210240530831,
+      "learning_rate": 9.91613191730907e-06,
+      "loss": 0.5692,
+      "step": 1360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6210779960478716,
+      "learning_rate": 9.915943662328555e-06,
+      "loss": 0.5621,
+      "step": 1361
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0351385027493207,
+      "learning_rate": 9.915755198092027e-06,
+      "loss": 0.6741,
+      "step": 1362
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0988988261684944,
+      "learning_rate": 9.915566524607509e-06,
+      "loss": 0.5733,
+      "step": 1363
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8624232098103166,
+      "learning_rate": 9.915377641883033e-06,
+      "loss": 0.6935,
+      "step": 1364
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9210306619681377,
+      "learning_rate": 9.915188549926637e-06,
+      "loss": 0.6424,
+      "step": 1365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9948571038404521,
+      "learning_rate": 9.914999248746371e-06,
+      "loss": 0.6696,
+      "step": 1366
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2880528637544018,
+      "learning_rate": 9.914809738350294e-06,
+      "loss": 0.6204,
+      "step": 1367
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6282329190313225,
+      "learning_rate": 9.914620018746472e-06,
+      "loss": 0.5488,
+      "step": 1368
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8313993414409726,
+      "learning_rate": 9.91443008994298e-06,
+      "loss": 0.6667,
+      "step": 1369
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1268742468420139,
+      "learning_rate": 9.914239951947902e-06,
+      "loss": 0.7016,
+      "step": 1370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7616004118200012,
+      "learning_rate": 9.914049604769335e-06,
+      "loss": 0.6171,
+      "step": 1371
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6443959873384107,
+      "learning_rate": 9.913859048415378e-06,
+      "loss": 0.5593,
+      "step": 1372
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8507864958821436,
+      "learning_rate": 9.913668282894144e-06,
+      "loss": 0.6136,
+      "step": 1373
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7798333503909397,
+      "learning_rate": 9.913477308213754e-06,
+      "loss": 0.6367,
+      "step": 1374
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8440171935341105,
+      "learning_rate": 9.913286124382335e-06,
+      "loss": 0.6558,
+      "step": 1375
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7562020603164891,
+      "learning_rate": 9.913094731408029e-06,
+      "loss": 0.6334,
+      "step": 1376
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7084724801967266,
+      "learning_rate": 9.912903129298977e-06,
+      "loss": 0.5323,
+      "step": 1377
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6243371740891822,
+      "learning_rate": 9.91271131806334e-06,
+      "loss": 0.5569,
+      "step": 1378
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5874709344697451,
+      "learning_rate": 9.91251929770928e-06,
+      "loss": 0.539,
+      "step": 1379
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.221005406389,
+      "learning_rate": 9.912327068244972e-06,
+      "loss": 0.6443,
+      "step": 1380
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6991482269482668,
+      "learning_rate": 9.912134629678598e-06,
+      "loss": 0.5285,
+      "step": 1381
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7167757674295981,
+      "learning_rate": 9.91194198201835e-06,
+      "loss": 0.5978,
+      "step": 1382
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5761426658547066,
+      "learning_rate": 9.911749125272428e-06,
+      "loss": 0.5181,
+      "step": 1383
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8663088232605922,
+      "learning_rate": 9.911556059449043e-06,
+      "loss": 0.6388,
+      "step": 1384
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.49120814305152,
+      "learning_rate": 9.91136278455641e-06,
+      "loss": 0.6325,
+      "step": 1385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.623522604963163,
+      "learning_rate": 9.911169300602758e-06,
+      "loss": 0.5337,
+      "step": 1386
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7734456232414336,
+      "learning_rate": 9.910975607596322e-06,
+      "loss": 0.5485,
+      "step": 1387
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5857910899105334,
+      "learning_rate": 9.91078170554535e-06,
+      "loss": 0.4878,
+      "step": 1388
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.089985469595313,
+      "learning_rate": 9.910587594458089e-06,
+      "loss": 0.6836,
+      "step": 1389
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8096160681242981,
+      "learning_rate": 9.910393274342811e-06,
+      "loss": 0.613,
+      "step": 1390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9494275055958991,
+      "learning_rate": 9.91019874520778e-06,
+      "loss": 0.6892,
+      "step": 1391
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8545086755605236,
+      "learning_rate": 9.910004007061281e-06,
+      "loss": 0.5985,
+      "step": 1392
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8613985857742912,
+      "learning_rate": 9.9098090599116e-06,
+      "loss": 0.6026,
+      "step": 1393
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6690755088707274,
+      "learning_rate": 9.909613903767038e-06,
+      "loss": 0.5481,
+      "step": 1394
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.649320401916616,
+      "learning_rate": 9.909418538635898e-06,
+      "loss": 0.5486,
+      "step": 1395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8956892661343855,
+      "learning_rate": 9.909222964526502e-06,
+      "loss": 0.6372,
+      "step": 1396
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6206099989106391,
+      "learning_rate": 9.909027181447171e-06,
+      "loss": 0.5523,
+      "step": 1397
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6666549049604876,
+      "learning_rate": 9.90883118940624e-06,
+      "loss": 0.4955,
+      "step": 1398
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9338525361206736,
+      "learning_rate": 9.908634988412052e-06,
+      "loss": 0.7067,
+      "step": 1399
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5636396475962866,
+      "learning_rate": 9.908438578472958e-06,
+      "loss": 0.502,
+      "step": 1400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9191880855606592,
+      "learning_rate": 9.908241959597317e-06,
+      "loss": 0.6488,
+      "step": 1401
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6686319641226389,
+      "learning_rate": 9.908045131793504e-06,
+      "loss": 0.533,
+      "step": 1402
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7011337591826701,
+      "learning_rate": 9.907848095069892e-06,
+      "loss": 0.5539,
+      "step": 1403
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6767613065949596,
+      "learning_rate": 9.907650849434868e-06,
+      "loss": 0.6041,
+      "step": 1404
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.832178458992157,
+      "learning_rate": 9.907453394896834e-06,
+      "loss": 0.6848,
+      "step": 1405
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7569351652390657,
+      "learning_rate": 9.907255731464186e-06,
+      "loss": 0.5541,
+      "step": 1406
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6697095708357677,
+      "learning_rate": 9.907057859145347e-06,
+      "loss": 0.4926,
+      "step": 1407
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6832580071791883,
+      "learning_rate": 9.906859777948735e-06,
+      "loss": 0.5754,
+      "step": 1408
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8892841362498346,
+      "learning_rate": 9.906661487882781e-06,
+      "loss": 0.6543,
+      "step": 1409
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6172957653929051,
+      "learning_rate": 9.906462988955927e-06,
+      "loss": 0.5229,
+      "step": 1410
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6165829961599874,
+      "learning_rate": 9.90626428117662e-06,
+      "loss": 0.5249,
+      "step": 1411
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9504490083975338,
+      "learning_rate": 9.906065364553325e-06,
+      "loss": 0.6192,
+      "step": 1412
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1873244845545143,
+      "learning_rate": 9.905866239094504e-06,
+      "loss": 0.6932,
+      "step": 1413
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.06923183666248,
+      "learning_rate": 9.905666904808634e-06,
+      "loss": 0.6468,
+      "step": 1414
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6611026336952993,
+      "learning_rate": 9.905467361704197e-06,
+      "loss": 0.5087,
+      "step": 1415
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8563940941522039,
+      "learning_rate": 9.905267609789694e-06,
+      "loss": 0.6809,
+      "step": 1416
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7925201345931134,
+      "learning_rate": 9.905067649073623e-06,
+      "loss": 0.577,
+      "step": 1417
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7296902809381323,
+      "learning_rate": 9.904867479564495e-06,
+      "loss": 0.5895,
+      "step": 1418
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7197783778450046,
+      "learning_rate": 9.904667101270832e-06,
+      "loss": 0.526,
+      "step": 1419
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9967702771570577,
+      "learning_rate": 9.904466514201166e-06,
+      "loss": 0.6622,
+      "step": 1420
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8584232773378019,
+      "learning_rate": 9.904265718364032e-06,
+      "loss": 0.5902,
+      "step": 1421
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0000577762329252,
+      "learning_rate": 9.904064713767978e-06,
+      "loss": 0.6025,
+      "step": 1422
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6847807358568969,
+      "learning_rate": 9.90386350042156e-06,
+      "loss": 0.5539,
+      "step": 1423
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0644584919556448,
+      "learning_rate": 9.903662078333342e-06,
+      "loss": 0.6682,
+      "step": 1424
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5823183939248158,
+      "learning_rate": 9.9034604475119e-06,
+      "loss": 0.5162,
+      "step": 1425
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9679891295378018,
+      "learning_rate": 9.903258607965818e-06,
+      "loss": 0.6741,
+      "step": 1426
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6281644095149793,
+      "learning_rate": 9.903056559703683e-06,
+      "loss": 0.5355,
+      "step": 1427
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9660733377233732,
+      "learning_rate": 9.902854302734099e-06,
+      "loss": 0.6062,
+      "step": 1428
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5278258123370565,
+      "learning_rate": 9.902651837065675e-06,
+      "loss": 0.6827,
+      "step": 1429
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8966393076686645,
+      "learning_rate": 9.902449162707029e-06,
+      "loss": 0.6557,
+      "step": 1430
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8878906994413838,
+      "learning_rate": 9.90224627966679e-06,
+      "loss": 0.6704,
+      "step": 1431
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1623132432739118,
+      "learning_rate": 9.902043187953589e-06,
+      "loss": 0.6952,
+      "step": 1432
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9283657657374793,
+      "learning_rate": 9.901839887576075e-06,
+      "loss": 0.6659,
+      "step": 1433
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9289844103113588,
+      "learning_rate": 9.901636378542902e-06,
+      "loss": 0.6226,
+      "step": 1434
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7697613872955419,
+      "learning_rate": 9.901432660862731e-06,
+      "loss": 0.5677,
+      "step": 1435
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8329828227111222,
+      "learning_rate": 9.901228734544235e-06,
+      "loss": 0.586,
+      "step": 1436
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6647831985745791,
+      "learning_rate": 9.901024599596092e-06,
+      "loss": 0.5618,
+      "step": 1437
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9703069379973013,
+      "learning_rate": 9.900820256026996e-06,
+      "loss": 0.6259,
+      "step": 1438
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7257311929494387,
+      "learning_rate": 9.90061570384564e-06,
+      "loss": 0.5675,
+      "step": 1439
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6434160744348719,
+      "learning_rate": 9.900410943060734e-06,
+      "loss": 0.5133,
+      "step": 1440
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.567375826023123,
+      "learning_rate": 9.900205973680996e-06,
+      "loss": 0.5638,
+      "step": 1441
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1463821681631006,
+      "learning_rate": 9.900000795715146e-06,
+      "loss": 0.6737,
+      "step": 1442
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7329192627325487,
+      "learning_rate": 9.899795409171923e-06,
+      "loss": 0.5904,
+      "step": 1443
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8684228605169513,
+      "learning_rate": 9.899589814060063e-06,
+      "loss": 0.627,
+      "step": 1444
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5880745445069139,
+      "learning_rate": 9.899384010388324e-06,
+      "loss": 0.5317,
+      "step": 1445
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6981588853474568,
+      "learning_rate": 9.899177998165464e-06,
+      "loss": 0.5199,
+      "step": 1446
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9437908832128978,
+      "learning_rate": 9.898971777400251e-06,
+      "loss": 0.6431,
+      "step": 1447
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0875796335565775,
+      "learning_rate": 9.898765348101464e-06,
+      "loss": 0.6638,
+      "step": 1448
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8013969374803733,
+      "learning_rate": 9.898558710277893e-06,
+      "loss": 0.6208,
+      "step": 1449
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.827640750552822,
+      "learning_rate": 9.898351863938328e-06,
+      "loss": 0.6842,
+      "step": 1450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8845593547754813,
+      "learning_rate": 9.898144809091578e-06,
+      "loss": 0.6881,
+      "step": 1451
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6339290565768775,
+      "learning_rate": 9.897937545746457e-06,
+      "loss": 0.5589,
+      "step": 1452
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.63683296212066,
+      "learning_rate": 9.897730073911785e-06,
+      "loss": 0.5677,
+      "step": 1453
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8037689871613029,
+      "learning_rate": 9.897522393596395e-06,
+      "loss": 0.6354,
+      "step": 1454
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7750823129055913,
+      "learning_rate": 9.897314504809128e-06,
+      "loss": 0.5568,
+      "step": 1455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6269252556078115,
+      "learning_rate": 9.89710640755883e-06,
+      "loss": 0.5016,
+      "step": 1456
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5762192182401308,
+      "learning_rate": 9.896898101854363e-06,
+      "loss": 0.5174,
+      "step": 1457
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5017342628673056,
+      "learning_rate": 9.896689587704591e-06,
+      "loss": 0.6123,
+      "step": 1458
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7616051248698723,
+      "learning_rate": 9.896480865118393e-06,
+      "loss": 0.5661,
+      "step": 1459
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6299426594954421,
+      "learning_rate": 9.896271934104649e-06,
+      "loss": 0.537,
+      "step": 1460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6590310849058559,
+      "learning_rate": 9.896062794672255e-06,
+      "loss": 0.5836,
+      "step": 1461
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6677798165719644,
+      "learning_rate": 9.895853446830115e-06,
+      "loss": 0.6128,
+      "step": 1462
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7311823577825591,
+      "learning_rate": 9.895643890587137e-06,
+      "loss": 0.6706,
+      "step": 1463
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5735339436796842,
+      "learning_rate": 9.895434125952244e-06,
+      "loss": 0.552,
+      "step": 1464
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6007134379056589,
+      "learning_rate": 9.895224152934362e-06,
+      "loss": 0.5615,
+      "step": 1465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.792092039855217,
+      "learning_rate": 9.895013971542433e-06,
+      "loss": 0.592,
+      "step": 1466
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9023231896386574,
+      "learning_rate": 9.8948035817854e-06,
+      "loss": 0.6443,
+      "step": 1467
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8971879997141177,
+      "learning_rate": 9.894592983672223e-06,
+      "loss": 0.6252,
+      "step": 1468
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6786774174016315,
+      "learning_rate": 9.89438217721186e-06,
+      "loss": 0.5836,
+      "step": 1469
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7149454501134153,
+      "learning_rate": 9.894171162413289e-06,
+      "loss": 0.5909,
+      "step": 1470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6392940942523266,
+      "learning_rate": 9.893959939285491e-06,
+      "loss": 0.5693,
+      "step": 1471
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8879975013720354,
+      "learning_rate": 9.893748507837458e-06,
+      "loss": 0.6327,
+      "step": 1472
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8588077828201054,
+      "learning_rate": 9.893536868078188e-06,
+      "loss": 0.6957,
+      "step": 1473
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9244754258118437,
+      "learning_rate": 9.893325020016692e-06,
+      "loss": 0.6857,
+      "step": 1474
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0370381276392335,
+      "learning_rate": 9.893112963661986e-06,
+      "loss": 0.5826,
+      "step": 1475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7727905979684659,
+      "learning_rate": 9.892900699023098e-06,
+      "loss": 0.6026,
+      "step": 1476
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6070696744331471,
+      "learning_rate": 9.892688226109064e-06,
+      "loss": 0.4797,
+      "step": 1477
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7962443280674872,
+      "learning_rate": 9.892475544928925e-06,
+      "loss": 0.5596,
+      "step": 1478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8647045792132472,
+      "learning_rate": 9.892262655491736e-06,
+      "loss": 0.6613,
+      "step": 1479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7776722259973866,
+      "learning_rate": 9.89204955780656e-06,
+      "loss": 0.6021,
+      "step": 1480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6041916236982071,
+      "learning_rate": 9.891836251882468e-06,
+      "loss": 0.4868,
+      "step": 1481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6806411366767511,
+      "learning_rate": 9.891622737728537e-06,
+      "loss": 0.5788,
+      "step": 1482
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7959127973674263,
+      "learning_rate": 9.891409015353859e-06,
+      "loss": 0.6046,
+      "step": 1483
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8228697951855963,
+      "learning_rate": 9.89119508476753e-06,
+      "loss": 0.5781,
+      "step": 1484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7064391469084955,
+      "learning_rate": 9.890980945978655e-06,
+      "loss": 0.5517,
+      "step": 1485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9005920425585042,
+      "learning_rate": 9.89076659899635e-06,
+      "loss": 0.6435,
+      "step": 1486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7604358015849504,
+      "learning_rate": 9.89055204382974e-06,
+      "loss": 0.591,
+      "step": 1487
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.684928912636846,
+      "learning_rate": 9.890337280487958e-06,
+      "loss": 0.5108,
+      "step": 1488
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8129477359373968,
+      "learning_rate": 9.890122308980145e-06,
+      "loss": 0.6252,
+      "step": 1489
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8150066125639978,
+      "learning_rate": 9.889907129315452e-06,
+      "loss": 0.6235,
+      "step": 1490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.902468709050217,
+      "learning_rate": 9.889691741503038e-06,
+      "loss": 0.6093,
+      "step": 1491
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.922199679349637,
+      "learning_rate": 9.889476145552073e-06,
+      "loss": 0.687,
+      "step": 1492
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7101187977843562,
+      "learning_rate": 9.889260341471732e-06,
+      "loss": 0.5222,
+      "step": 1493
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6516247339669807,
+      "learning_rate": 9.889044329271202e-06,
+      "loss": 0.6044,
+      "step": 1494
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6617508541540921,
+      "learning_rate": 9.888828108959678e-06,
+      "loss": 0.5943,
+      "step": 1495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6547760173649779,
+      "learning_rate": 9.888611680546366e-06,
+      "loss": 0.5926,
+      "step": 1496
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8146903408494094,
+      "learning_rate": 9.888395044040475e-06,
+      "loss": 0.6898,
+      "step": 1497
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9730046667110773,
+      "learning_rate": 9.888178199451227e-06,
+      "loss": 0.6485,
+      "step": 1498
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5776389634461832,
+      "learning_rate": 9.887961146787854e-06,
+      "loss": 0.5667,
+      "step": 1499
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9170097770006553,
+      "learning_rate": 9.887743886059595e-06,
+      "loss": 0.6274,
+      "step": 1500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.675811892876071,
+      "learning_rate": 9.887526417275698e-06,
+      "loss": 0.5421,
+      "step": 1501
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8182205446623694,
+      "learning_rate": 9.88730874044542e-06,
+      "loss": 0.6506,
+      "step": 1502
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.688221644284625,
+      "learning_rate": 9.887090855578026e-06,
+      "loss": 0.5645,
+      "step": 1503
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7839834339247176,
+      "learning_rate": 9.886872762682791e-06,
+      "loss": 0.6135,
+      "step": 1504
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7414209322009344,
+      "learning_rate": 9.886654461768999e-06,
+      "loss": 0.5907,
+      "step": 1505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7005784343852325,
+      "learning_rate": 9.886435952845941e-06,
+      "loss": 0.5969,
+      "step": 1506
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9034003306411155,
+      "learning_rate": 9.886217235922921e-06,
+      "loss": 0.6325,
+      "step": 1507
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6488541912557233,
+      "learning_rate": 9.885998311009246e-06,
+      "loss": 0.5933,
+      "step": 1508
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5939210241457038,
+      "learning_rate": 9.885779178114236e-06,
+      "loss": 0.5055,
+      "step": 1509
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7107377608775156,
+      "learning_rate": 9.88555983724722e-06,
+      "loss": 0.5928,
+      "step": 1510
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1112842519856718,
+      "learning_rate": 9.885340288417535e-06,
+      "loss": 0.7156,
+      "step": 1511
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6305165668223054,
+      "learning_rate": 9.885120531634525e-06,
+      "loss": 0.614,
+      "step": 1512
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7158541089272958,
+      "learning_rate": 9.884900566907543e-06,
+      "loss": 0.5829,
+      "step": 1513
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9801514047422134,
+      "learning_rate": 9.884680394245953e-06,
+      "loss": 0.6375,
+      "step": 1514
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8273363649517759,
+      "learning_rate": 9.88446001365913e-06,
+      "loss": 0.6249,
+      "step": 1515
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7454375934919218,
+      "learning_rate": 9.884239425156452e-06,
+      "loss": 0.6193,
+      "step": 1516
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7092783240013331,
+      "learning_rate": 9.88401862874731e-06,
+      "loss": 0.5809,
+      "step": 1517
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6810210577340193,
+      "learning_rate": 9.8837976244411e-06,
+      "loss": 0.5678,
+      "step": 1518
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6809892291975016,
+      "learning_rate": 9.883576412247233e-06,
+      "loss": 0.5781,
+      "step": 1519
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8503934579940726,
+      "learning_rate": 9.883354992175124e-06,
+      "loss": 0.6795,
+      "step": 1520
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9274211197639796,
+      "learning_rate": 9.883133364234198e-06,
+      "loss": 0.7125,
+      "step": 1521
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9042072813027953,
+      "learning_rate": 9.88291152843389e-06,
+      "loss": 0.7166,
+      "step": 1522
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6546267571881859,
+      "learning_rate": 9.88268948478364e-06,
+      "loss": 0.5698,
+      "step": 1523
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6648143273555109,
+      "learning_rate": 9.882467233292901e-06,
+      "loss": 0.6095,
+      "step": 1524
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.669896902749318,
+      "learning_rate": 9.882244773971137e-06,
+      "loss": 0.6117,
+      "step": 1525
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.669379190520074,
+      "learning_rate": 9.88202210682781e-06,
+      "loss": 0.5603,
+      "step": 1526
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6705263480036763,
+      "learning_rate": 9.881799231872406e-06,
+      "loss": 0.5099,
+      "step": 1527
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5993458923502961,
+      "learning_rate": 9.881576149114407e-06,
+      "loss": 0.509,
+      "step": 1528
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0238276485462094,
+      "learning_rate": 9.88135285856331e-06,
+      "loss": 0.5775,
+      "step": 1529
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8936331822670118,
+      "learning_rate": 9.881129360228623e-06,
+      "loss": 0.6531,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9032830182022972,
+      "learning_rate": 9.880905654119856e-06,
+      "loss": 0.5964,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8753903201065488,
+      "learning_rate": 9.880681740246531e-06,
+      "loss": 0.6343,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7976664177553304,
+      "learning_rate": 9.880457618618184e-06,
+      "loss": 0.558,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.769384373440501,
+      "learning_rate": 9.88023328924435e-06,
+      "loss": 0.5724,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.708971831517012,
+      "learning_rate": 9.880008752134578e-06,
+      "loss": 0.5605,
+      "step": 1535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6838771857352796,
+      "learning_rate": 9.87978400729843e-06,
+      "loss": 0.5889,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7179981698375474,
+      "learning_rate": 9.87955905474547e-06,
+      "loss": 0.6386,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7286557985737631,
+      "learning_rate": 9.879333894485272e-06,
+      "loss": 0.5959,
+      "step": 1538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6417993377828476,
+      "learning_rate": 9.879108526527425e-06,
+      "loss": 0.5761,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6455887264965151,
+      "learning_rate": 9.878882950881518e-06,
+      "loss": 0.5701,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9392531020535543,
+      "learning_rate": 9.878657167557156e-06,
+      "loss": 0.6737,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8052631006205957,
+      "learning_rate": 9.878431176563945e-06,
+      "loss": 0.6034,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6452744052517272,
+      "learning_rate": 9.87820497791151e-06,
+      "loss": 0.5617,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8888982388016089,
+      "learning_rate": 9.877978571609479e-06,
+      "loss": 0.6541,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8622303599846088,
+      "learning_rate": 9.877751957667486e-06,
+      "loss": 0.6467,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.760571615055131,
+      "learning_rate": 9.87752513609518e-06,
+      "loss": 0.615,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6745613351888998,
+      "learning_rate": 9.877298106902216e-06,
+      "loss": 0.5902,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6051554949691383,
+      "learning_rate": 9.877070870098256e-06,
+      "loss": 0.5162,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6335179298694252,
+      "learning_rate": 9.876843425692975e-06,
+      "loss": 0.5526,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8044564719001647,
+      "learning_rate": 9.876615773696053e-06,
+      "loss": 0.6415,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2969363633187778,
+      "learning_rate": 9.876387914117181e-06,
+      "loss": 0.6653,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8054796427768358,
+      "learning_rate": 9.876159846966057e-06,
+      "loss": 0.5273,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9842086078093535,
+      "learning_rate": 9.87593157225239e-06,
+      "loss": 0.5805,
+      "step": 1553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6891899317152046,
+      "learning_rate": 9.8757030899859e-06,
+      "loss": 0.5981,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3612018114252962,
+      "learning_rate": 9.875474400176307e-06,
+      "loss": 0.6092,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6543286888948006,
+      "learning_rate": 9.87524550283335e-06,
+      "loss": 0.5868,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.769976874425928,
+      "learning_rate": 9.875016397966772e-06,
+      "loss": 0.6744,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.585257672773087,
+      "learning_rate": 9.874787085586323e-06,
+      "loss": 0.6127,
+      "step": 1558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5717036765553271,
+      "learning_rate": 9.874557565701766e-06,
+      "loss": 0.5382,
+      "step": 1559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6541689594121229,
+      "learning_rate": 9.87432783832287e-06,
+      "loss": 0.5523,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8055566439494736,
+      "learning_rate": 9.874097903459413e-06,
+      "loss": 0.6251,
+      "step": 1561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8346090143319688,
+      "learning_rate": 9.873867761121186e-06,
+      "loss": 0.6496,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.670308442840768,
+      "learning_rate": 9.873637411317982e-06,
+      "loss": 0.5613,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6667886214579567,
+      "learning_rate": 9.873406854059606e-06,
+      "loss": 0.5542,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6787516379371499,
+      "learning_rate": 9.873176089355876e-06,
+      "loss": 0.5624,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7509925927710046,
+      "learning_rate": 9.87294511721661e-06,
+      "loss": 0.5894,
+      "step": 1566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9022552804982648,
+      "learning_rate": 9.872713937651644e-06,
+      "loss": 0.6462,
+      "step": 1567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6648974660353939,
+      "learning_rate": 9.872482550670815e-06,
+      "loss": 0.6236,
+      "step": 1568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9422739387844301,
+      "learning_rate": 9.872250956283974e-06,
+      "loss": 0.595,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7314958690611608,
+      "learning_rate": 9.87201915450098e-06,
+      "loss": 0.6098,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7636374468141408,
+      "learning_rate": 9.8717871453317e-06,
+      "loss": 0.6245,
+      "step": 1571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8121591259993941,
+      "learning_rate": 9.871554928786007e-06,
+      "loss": 0.5783,
+      "step": 1572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9954410395816852,
+      "learning_rate": 9.87132250487379e-06,
+      "loss": 0.6484,
+      "step": 1573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7954625985334292,
+      "learning_rate": 9.87108987360494e-06,
+      "loss": 0.5603,
+      "step": 1574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9622292414595693,
+      "learning_rate": 9.870857034989359e-06,
+      "loss": 0.6343,
+      "step": 1575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6513245173343255,
+      "learning_rate": 9.87062398903696e-06,
+      "loss": 0.554,
+      "step": 1576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7017719471811728,
+      "learning_rate": 9.87039073575766e-06,
+      "loss": 0.5777,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8207537650088288,
+      "learning_rate": 9.870157275161394e-06,
+      "loss": 0.5658,
+      "step": 1578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9306911535789684,
+      "learning_rate": 9.869923607258093e-06,
+      "loss": 0.5868,
+      "step": 1579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8082491365487466,
+      "learning_rate": 9.869689732057706e-06,
+      "loss": 0.7051,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7968390618902453,
+      "learning_rate": 9.869455649570188e-06,
+      "loss": 0.6241,
+      "step": 1581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7852533651109809,
+      "learning_rate": 9.869221359805506e-06,
+      "loss": 0.6005,
+      "step": 1582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.844980110728931,
+      "learning_rate": 9.868986862773628e-06,
+      "loss": 0.6541,
+      "step": 1583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.890366904680627,
+      "learning_rate": 9.86875215848454e-06,
+      "loss": 0.6543,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6280073463676458,
+      "learning_rate": 9.868517246948228e-06,
+      "loss": 0.5546,
+      "step": 1585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.682946858818819,
+      "learning_rate": 9.868282128174698e-06,
+      "loss": 0.5619,
+      "step": 1586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6369580171058496,
+      "learning_rate": 9.868046802173952e-06,
+      "loss": 0.5563,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6459659497595663,
+      "learning_rate": 9.867811268956011e-06,
+      "loss": 0.5679,
+      "step": 1588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9076663467022951,
+      "learning_rate": 9.867575528530899e-06,
+      "loss": 0.6413,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7644414635666698,
+      "learning_rate": 9.867339580908652e-06,
+      "loss": 0.5695,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9802070894770096,
+      "learning_rate": 9.867103426099313e-06,
+      "loss": 0.6452,
+      "step": 1591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9292906996900488,
+      "learning_rate": 9.866867064112934e-06,
+      "loss": 0.6407,
+      "step": 1592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6791333305841595,
+      "learning_rate": 9.866630494959574e-06,
+      "loss": 0.5767,
+      "step": 1593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6453810795437466,
+      "learning_rate": 9.866393718649309e-06,
+      "loss": 0.5665,
+      "step": 1594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8320577331876663,
+      "learning_rate": 9.866156735192211e-06,
+      "loss": 0.6371,
+      "step": 1595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7034813881171303,
+      "learning_rate": 9.865919544598373e-06,
+      "loss": 0.5716,
+      "step": 1596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7509585379589557,
+      "learning_rate": 9.865682146877888e-06,
+      "loss": 0.5962,
+      "step": 1597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9686834691796042,
+      "learning_rate": 9.865444542040865e-06,
+      "loss": 0.6133,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8268502522136598,
+      "learning_rate": 9.865206730097413e-06,
+      "loss": 0.6333,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1796734025912674,
+      "learning_rate": 9.86496871105766e-06,
+      "loss": 0.6021,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5929343695552657,
+      "learning_rate": 9.86473048493173e-06,
+      "loss": 0.4875,
+      "step": 1601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8711210540344185,
+      "learning_rate": 9.864492051729772e-06,
+      "loss": 0.6851,
+      "step": 1602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9756325623444182,
+      "learning_rate": 9.864253411461932e-06,
+      "loss": 0.6592,
+      "step": 1603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8278114941762432,
+      "learning_rate": 9.864014564138369e-06,
+      "loss": 0.6452,
+      "step": 1604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8427554962800573,
+      "learning_rate": 9.86377550976925e-06,
+      "loss": 0.6274,
+      "step": 1605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.780829481157279,
+      "learning_rate": 9.863536248364748e-06,
+      "loss": 0.6168,
+      "step": 1606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8066292955098436,
+      "learning_rate": 9.86329677993505e-06,
+      "loss": 0.5908,
+      "step": 1607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5680189296272247,
+      "learning_rate": 9.863057104490348e-06,
+      "loss": 0.5663,
+      "step": 1608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8443458599638527,
+      "learning_rate": 9.862817222040845e-06,
+      "loss": 0.6007,
+      "step": 1609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9003249014982616,
+      "learning_rate": 9.862577132596755e-06,
+      "loss": 0.6193,
+      "step": 1610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.852654164152259,
+      "learning_rate": 9.862336836168293e-06,
+      "loss": 0.6172,
+      "step": 1611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5795560644283292,
+      "learning_rate": 9.86209633276569e-06,
+      "loss": 0.547,
+      "step": 1612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6376658582526519,
+      "learning_rate": 9.861855622399184e-06,
+      "loss": 0.5538,
+      "step": 1613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6823015338001743,
+      "learning_rate": 9.86161470507902e-06,
+      "loss": 0.5244,
+      "step": 1614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7020896797015818,
+      "learning_rate": 9.861373580815452e-06,
+      "loss": 0.6004,
+      "step": 1615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7067135748192178,
+      "learning_rate": 9.861132249618745e-06,
+      "loss": 0.5495,
+      "step": 1616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7840787137883893,
+      "learning_rate": 9.860890711499175e-06,
+      "loss": 0.6569,
+      "step": 1617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6508566226300785,
+      "learning_rate": 9.860648966467018e-06,
+      "loss": 0.5679,
+      "step": 1618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7997398853337984,
+      "learning_rate": 9.860407014532568e-06,
+      "loss": 0.63,
+      "step": 1619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6305162767609661,
+      "learning_rate": 9.860164855706123e-06,
+      "loss": 0.5793,
+      "step": 1620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6898611702327054,
+      "learning_rate": 9.859922489997991e-06,
+      "loss": 0.5842,
+      "step": 1621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6204437929878805,
+      "learning_rate": 9.859679917418489e-06,
+      "loss": 0.5621,
+      "step": 1622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8567548920555998,
+      "learning_rate": 9.859437137977942e-06,
+      "loss": 0.6206,
+      "step": 1623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6573511487962571,
+      "learning_rate": 9.859194151686685e-06,
+      "loss": 0.5791,
+      "step": 1624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7905473570590201,
+      "learning_rate": 9.85895095855506e-06,
+      "loss": 0.5781,
+      "step": 1625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9481362480472231,
+      "learning_rate": 9.858707558593421e-06,
+      "loss": 0.6792,
+      "step": 1626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.838869534675686,
+      "learning_rate": 9.858463951812125e-06,
+      "loss": 0.6789,
+      "step": 1627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6271821937311737,
+      "learning_rate": 9.858220138221546e-06,
+      "loss": 0.531,
+      "step": 1628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6543623055698378,
+      "learning_rate": 9.85797611783206e-06,
+      "loss": 0.5008,
+      "step": 1629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7101443776889375,
+      "learning_rate": 9.857731890654055e-06,
+      "loss": 0.5595,
+      "step": 1630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.933466387973217,
+      "learning_rate": 9.857487456697925e-06,
+      "loss": 0.576,
+      "step": 1631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.408488406231972,
+      "learning_rate": 9.857242815974078e-06,
+      "loss": 0.6954,
+      "step": 1632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9322296346854334,
+      "learning_rate": 9.856997968492924e-06,
+      "loss": 0.6195,
+      "step": 1633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8527005341614082,
+      "learning_rate": 9.856752914264889e-06,
+      "loss": 0.6158,
+      "step": 1634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6029817931915498,
+      "learning_rate": 9.8565076533004e-06,
+      "loss": 0.5773,
+      "step": 1635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6922544203961906,
+      "learning_rate": 9.856262185609901e-06,
+      "loss": 0.5389,
+      "step": 1636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6657983220041698,
+      "learning_rate": 9.856016511203839e-06,
+      "loss": 0.5861,
+      "step": 1637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0928942883190693,
+      "learning_rate": 9.855770630092672e-06,
+      "loss": 0.6335,
+      "step": 1638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1189330925779744,
+      "learning_rate": 9.855524542286866e-06,
+      "loss": 0.606,
+      "step": 1639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8953629009058321,
+      "learning_rate": 9.855278247796897e-06,
+      "loss": 0.5854,
+      "step": 1640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6875087356327355,
+      "learning_rate": 9.855031746633247e-06,
+      "loss": 0.5434,
+      "step": 1641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1680271686982153,
+      "learning_rate": 9.854785038806411e-06,
+      "loss": 0.65,
+      "step": 1642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6489497901061975,
+      "learning_rate": 9.854538124326889e-06,
+      "loss": 0.6051,
+      "step": 1643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6195205120566304,
+      "learning_rate": 9.854291003205192e-06,
+      "loss": 0.5157,
+      "step": 1644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8719980440711714,
+      "learning_rate": 9.854043675451838e-06,
+      "loss": 0.6223,
+      "step": 1645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8184096554586712,
+      "learning_rate": 9.853796141077357e-06,
+      "loss": 0.6516,
+      "step": 1646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6511274776408341,
+      "learning_rate": 9.853548400092285e-06,
+      "loss": 0.5093,
+      "step": 1647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9905158951835199,
+      "learning_rate": 9.853300452507167e-06,
+      "loss": 0.6101,
+      "step": 1648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6572766198066207,
+      "learning_rate": 9.853052298332559e-06,
+      "loss": 0.5691,
+      "step": 1649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6855701865032515,
+      "learning_rate": 9.85280393757902e-06,
+      "loss": 0.541,
+      "step": 1650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7668807561188197,
+      "learning_rate": 9.852555370257127e-06,
+      "loss": 0.6382,
+      "step": 1651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6325259906279111,
+      "learning_rate": 9.852306596377459e-06,
+      "loss": 0.5566,
+      "step": 1652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9250753695471651,
+      "learning_rate": 9.852057615950603e-06,
+      "loss": 0.6393,
+      "step": 1653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9279146497246463,
+      "learning_rate": 9.85180842898716e-06,
+      "loss": 0.6717,
+      "step": 1654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7987768441302029,
+      "learning_rate": 9.851559035497735e-06,
+      "loss": 0.6624,
+      "step": 1655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5686685449129294,
+      "learning_rate": 9.851309435492948e-06,
+      "loss": 0.5228,
+      "step": 1656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7910938429868714,
+      "learning_rate": 9.85105962898342e-06,
+      "loss": 0.6313,
+      "step": 1657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7554652092727698,
+      "learning_rate": 9.850809615979785e-06,
+      "loss": 0.6802,
+      "step": 1658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9099323233798392,
+      "learning_rate": 9.850559396492685e-06,
+      "loss": 0.655,
+      "step": 1659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8393170889094593,
+      "learning_rate": 9.850308970532772e-06,
+      "loss": 0.6576,
+      "step": 1660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7841118211705639,
+      "learning_rate": 9.850058338110705e-06,
+      "loss": 0.5871,
+      "step": 1661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5549050394204721,
+      "learning_rate": 9.849807499237154e-06,
+      "loss": 0.5044,
+      "step": 1662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.686561215272153,
+      "learning_rate": 9.849556453922796e-06,
+      "loss": 0.5421,
+      "step": 1663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5762718732077279,
+      "learning_rate": 9.849305202178314e-06,
+      "loss": 0.5529,
+      "step": 1664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.758757336304438,
+      "learning_rate": 9.849053744014408e-06,
+      "loss": 0.6206,
+      "step": 1665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.55247077233078,
+      "learning_rate": 9.848802079441779e-06,
+      "loss": 0.5167,
+      "step": 1666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8895213823591999,
+      "learning_rate": 9.848550208471142e-06,
+      "loss": 0.6165,
+      "step": 1667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.855374858594082,
+      "learning_rate": 9.848298131113214e-06,
+      "loss": 0.6212,
+      "step": 1668
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6102036648610574,
+      "learning_rate": 9.848045847378726e-06,
+      "loss": 0.5034,
+      "step": 1669
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.825259433810369,
+      "learning_rate": 9.847793357278421e-06,
+      "loss": 0.6835,
+      "step": 1670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6130078610734073,
+      "learning_rate": 9.847540660823045e-06,
+      "loss": 0.5656,
+      "step": 1671
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7704459089579551,
+      "learning_rate": 9.847287758023351e-06,
+      "loss": 0.6029,
+      "step": 1672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6260214451166097,
+      "learning_rate": 9.847034648890108e-06,
+      "loss": 0.5277,
+      "step": 1673
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8269750454306841,
+      "learning_rate": 9.84678133343409e-06,
+      "loss": 0.5882,
+      "step": 1674
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7715515416702583,
+      "learning_rate": 9.846527811666078e-06,
+      "loss": 0.5867,
+      "step": 1675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6165145422690684,
+      "learning_rate": 9.846274083596863e-06,
+      "loss": 0.5475,
+      "step": 1676
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6233397616203508,
+      "learning_rate": 9.846020149237248e-06,
+      "loss": 0.545,
+      "step": 1677
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6156061510869181,
+      "learning_rate": 9.845766008598042e-06,
+      "loss": 0.5755,
+      "step": 1678
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9484201167773875,
+      "learning_rate": 9.845511661690059e-06,
+      "loss": 0.6152,
+      "step": 1679
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6501776492838833,
+      "learning_rate": 9.84525710852413e-06,
+      "loss": 0.5681,
+      "step": 1680
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6109494161031894,
+      "learning_rate": 9.84500234911109e-06,
+      "loss": 0.5359,
+      "step": 1681
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6539827487484311,
+      "learning_rate": 9.84474738346178e-06,
+      "loss": 0.5798,
+      "step": 1682
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.634957277775411,
+      "learning_rate": 9.844492211587058e-06,
+      "loss": 0.5734,
+      "step": 1683
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6636185620154464,
+      "learning_rate": 9.844236833497782e-06,
+      "loss": 0.5195,
+      "step": 1684
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2841736516006863,
+      "learning_rate": 9.843981249204825e-06,
+      "loss": 0.6146,
+      "step": 1685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7380799101894738,
+      "learning_rate": 9.843725458719065e-06,
+      "loss": 0.5537,
+      "step": 1686
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6474222869333314,
+      "learning_rate": 9.84346946205139e-06,
+      "loss": 0.5762,
+      "step": 1687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6328728625853145,
+      "learning_rate": 9.843213259212698e-06,
+      "loss": 0.5034,
+      "step": 1688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9351332256119143,
+      "learning_rate": 9.842956850213893e-06,
+      "loss": 0.6256,
+      "step": 1689
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6733164129517442,
+      "learning_rate": 9.842700235065893e-06,
+      "loss": 0.6195,
+      "step": 1690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5682114661543541,
+      "learning_rate": 9.842443413779618e-06,
+      "loss": 0.5294,
+      "step": 1691
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6413366655094044,
+      "learning_rate": 9.842186386366002e-06,
+      "loss": 0.5897,
+      "step": 1692
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.63052655305629,
+      "learning_rate": 9.841929152835983e-06,
+      "loss": 0.5714,
+      "step": 1693
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.67460003471768,
+      "learning_rate": 9.841671713200513e-06,
+      "loss": 0.548,
+      "step": 1694
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8176681406211385,
+      "learning_rate": 9.841414067470552e-06,
+      "loss": 0.6159,
+      "step": 1695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9175096301150686,
+      "learning_rate": 9.841156215657063e-06,
+      "loss": 0.6853,
+      "step": 1696
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7801575625897028,
+      "learning_rate": 9.840898157771024e-06,
+      "loss": 0.5602,
+      "step": 1697
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6269357761719352,
+      "learning_rate": 9.84063989382342e-06,
+      "loss": 0.5295,
+      "step": 1698
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7248816894100609,
+      "learning_rate": 9.840381423825245e-06,
+      "loss": 0.6128,
+      "step": 1699
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8856359618494074,
+      "learning_rate": 9.840122747787499e-06,
+      "loss": 0.6154,
+      "step": 1700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8144981201588504,
+      "learning_rate": 9.839863865721197e-06,
+      "loss": 0.6517,
+      "step": 1701
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8074339626714965,
+      "learning_rate": 9.839604777637355e-06,
+      "loss": 0.6454,
+      "step": 1702
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.755912866391722,
+      "learning_rate": 9.839345483547002e-06,
+      "loss": 0.5942,
+      "step": 1703
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5453412409044164,
+      "learning_rate": 9.839085983461176e-06,
+      "loss": 0.4867,
+      "step": 1704
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1312225476673403,
+      "learning_rate": 9.838826277390924e-06,
+      "loss": 0.623,
+      "step": 1705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5580077144699175,
+      "learning_rate": 9.8385663653473e-06,
+      "loss": 0.5262,
+      "step": 1706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0276120472248955,
+      "learning_rate": 9.838306247341368e-06,
+      "loss": 0.6897,
+      "step": 1707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6345644554063219,
+      "learning_rate": 9.838045923384198e-06,
+      "loss": 0.5501,
+      "step": 1708
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6950163591199844,
+      "learning_rate": 9.837785393486875e-06,
+      "loss": 0.5713,
+      "step": 1709
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7247758833337077,
+      "learning_rate": 9.837524657660486e-06,
+      "loss": 0.6012,
+      "step": 1710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8438527231007332,
+      "learning_rate": 9.837263715916132e-06,
+      "loss": 0.6684,
+      "step": 1711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.003086199282932,
+      "learning_rate": 9.837002568264919e-06,
+      "loss": 0.5959,
+      "step": 1712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9463278570698767,
+      "learning_rate": 9.836741214717964e-06,
+      "loss": 0.6049,
+      "step": 1713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6365549845329462,
+      "learning_rate": 9.836479655286391e-06,
+      "loss": 0.5541,
+      "step": 1714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8799062000288134,
+      "learning_rate": 9.836217889981335e-06,
+      "loss": 0.6678,
+      "step": 1715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6781532713920536,
+      "learning_rate": 9.835955918813937e-06,
+      "loss": 0.554,
+      "step": 1716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6090068756136409,
+      "learning_rate": 9.835693741795348e-06,
+      "loss": 0.5387,
+      "step": 1717
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6727395605649175,
+      "learning_rate": 9.83543135893673e-06,
+      "loss": 0.5264,
+      "step": 1718
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7648070527881605,
+      "learning_rate": 9.83516877024925e-06,
+      "loss": 0.6062,
+      "step": 1719
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8421626479733515,
+      "learning_rate": 9.83490597574409e-06,
+      "loss": 0.5994,
+      "step": 1720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.742540896777215,
+      "learning_rate": 9.83464297543243e-06,
+      "loss": 0.5347,
+      "step": 1721
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8461047171388711,
+      "learning_rate": 9.834379769325466e-06,
+      "loss": 0.6343,
+      "step": 1722
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6598525582260091,
+      "learning_rate": 9.834116357434407e-06,
+      "loss": 0.59,
+      "step": 1723
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6924863493852103,
+      "learning_rate": 9.833852739770463e-06,
+      "loss": 0.5899,
+      "step": 1724
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6556261440230574,
+      "learning_rate": 9.833588916344851e-06,
+      "loss": 0.5352,
+      "step": 1725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6164402957827337,
+      "learning_rate": 9.833324887168807e-06,
+      "loss": 0.5855,
+      "step": 1726
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.83313658454367,
+      "learning_rate": 9.833060652253567e-06,
+      "loss": 0.6342,
+      "step": 1727
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7971140584665952,
+      "learning_rate": 9.83279621161038e-06,
+      "loss": 0.6277,
+      "step": 1728
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7533113921604073,
+      "learning_rate": 9.832531565250501e-06,
+      "loss": 0.5854,
+      "step": 1729
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5973559185082606,
+      "learning_rate": 9.832266713185195e-06,
+      "loss": 0.5474,
+      "step": 1730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6941911837247505,
+      "learning_rate": 9.832001655425737e-06,
+      "loss": 0.5791,
+      "step": 1731
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7588200922371693,
+      "learning_rate": 9.83173639198341e-06,
+      "loss": 0.5659,
+      "step": 1732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.659717998884332,
+      "learning_rate": 9.831470922869506e-06,
+      "loss": 0.5656,
+      "step": 1733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.792657017702605,
+      "learning_rate": 9.831205248095322e-06,
+      "loss": 0.5704,
+      "step": 1734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8794596488407359,
+      "learning_rate": 9.83093936767217e-06,
+      "loss": 0.5796,
+      "step": 1735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6368550476640701,
+      "learning_rate": 9.830673281611367e-06,
+      "loss": 0.5795,
+      "step": 1736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8113624566405523,
+      "learning_rate": 9.830406989924239e-06,
+      "loss": 0.664,
+      "step": 1737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7148748320239768,
+      "learning_rate": 9.83014049262212e-06,
+      "loss": 0.5796,
+      "step": 1738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7045937711178463,
+      "learning_rate": 9.829873789716355e-06,
+      "loss": 0.5847,
+      "step": 1739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5964086549917872,
+      "learning_rate": 9.829606881218297e-06,
+      "loss": 0.5836,
+      "step": 1740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8052988769986834,
+      "learning_rate": 9.829339767139308e-06,
+      "loss": 0.5969,
+      "step": 1741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9221403141530807,
+      "learning_rate": 9.82907244749076e-06,
+      "loss": 0.7186,
+      "step": 1742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6203895913139273,
+      "learning_rate": 9.828804922284026e-06,
+      "loss": 0.5362,
+      "step": 1743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8738324041496631,
+      "learning_rate": 9.828537191530496e-06,
+      "loss": 0.614,
+      "step": 1744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7098802141428318,
+      "learning_rate": 9.82826925524157e-06,
+      "loss": 0.6196,
+      "step": 1745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6802253944524707,
+      "learning_rate": 9.82800111342865e-06,
+      "loss": 0.5573,
+      "step": 1746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8081654337036757,
+      "learning_rate": 9.82773276610315e-06,
+      "loss": 0.6616,
+      "step": 1747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7162665211569121,
+      "learning_rate": 9.827464213276494e-06,
+      "loss": 0.5565,
+      "step": 1748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2342725550537075,
+      "learning_rate": 9.827195454960115e-06,
+      "loss": 0.6428,
+      "step": 1749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8406360594694418,
+      "learning_rate": 9.826926491165448e-06,
+      "loss": 0.6361,
+      "step": 1750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7423596412830483,
+      "learning_rate": 9.826657321903946e-06,
+      "loss": 0.6112,
+      "step": 1751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6104102388515092,
+      "learning_rate": 9.826387947187066e-06,
+      "loss": 0.516,
+      "step": 1752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6846252071575041,
+      "learning_rate": 9.826118367026274e-06,
+      "loss": 0.5526,
+      "step": 1753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8553823260980193,
+      "learning_rate": 9.825848581433044e-06,
+      "loss": 0.704,
+      "step": 1754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6719827243923429,
+      "learning_rate": 9.825578590418862e-06,
+      "loss": 0.5003,
+      "step": 1755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7850911675406774,
+      "learning_rate": 9.82530839399522e-06,
+      "loss": 0.5521,
+      "step": 1756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7978646385458676,
+      "learning_rate": 9.825037992173618e-06,
+      "loss": 0.6446,
+      "step": 1757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8090192887152099,
+      "learning_rate": 9.824767384965567e-06,
+      "loss": 0.6247,
+      "step": 1758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8957108224116083,
+      "learning_rate": 9.82449657238259e-06,
+      "loss": 0.6446,
+      "step": 1759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8677758678006399,
+      "learning_rate": 9.824225554436207e-06,
+      "loss": 0.6161,
+      "step": 1760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8454027020819754,
+      "learning_rate": 9.82395433113796e-06,
+      "loss": 0.6067,
+      "step": 1761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7106692070370019,
+      "learning_rate": 9.82368290249939e-06,
+      "loss": 0.6342,
+      "step": 1762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6807978308051146,
+      "learning_rate": 9.823411268532055e-06,
+      "loss": 0.5495,
+      "step": 1763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6558366754497376,
+      "learning_rate": 9.823139429247515e-06,
+      "loss": 0.5589,
+      "step": 1764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7010054377950301,
+      "learning_rate": 9.822867384657341e-06,
+      "loss": 0.5249,
+      "step": 1765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6448064571642647,
+      "learning_rate": 9.822595134773116e-06,
+      "loss": 0.5513,
+      "step": 1766
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0398190136124312,
+      "learning_rate": 9.822322679606427e-06,
+      "loss": 0.6578,
+      "step": 1767
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6064319244857759,
+      "learning_rate": 9.822050019168871e-06,
+      "loss": 0.5589,
+      "step": 1768
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7163008061551831,
+      "learning_rate": 9.821777153472053e-06,
+      "loss": 0.5687,
+      "step": 1769
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6828781593595651,
+      "learning_rate": 9.821504082527592e-06,
+      "loss": 0.5684,
+      "step": 1770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7278664732198938,
+      "learning_rate": 9.821230806347111e-06,
+      "loss": 0.5327,
+      "step": 1771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7243893679333716,
+      "learning_rate": 9.820957324942237e-06,
+      "loss": 0.5791,
+      "step": 1772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.764962784739967,
+      "learning_rate": 9.820683638324618e-06,
+      "loss": 0.5995,
+      "step": 1773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1084149493792808,
+      "learning_rate": 9.820409746505901e-06,
+      "loss": 0.6086,
+      "step": 1774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6267456203921593,
+      "learning_rate": 9.820135649497744e-06,
+      "loss": 0.5813,
+      "step": 1775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9700337915284244,
+      "learning_rate": 9.819861347311815e-06,
+      "loss": 0.6304,
+      "step": 1776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9323486483003622,
+      "learning_rate": 9.819586839959792e-06,
+      "loss": 0.6092,
+      "step": 1777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8293205415777264,
+      "learning_rate": 9.819312127453358e-06,
+      "loss": 0.6256,
+      "step": 1778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7672079298631141,
+      "learning_rate": 9.819037209804205e-06,
+      "loss": 0.6065,
+      "step": 1779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6515826109273735,
+      "learning_rate": 9.818762087024039e-06,
+      "loss": 0.5435,
+      "step": 1780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6496504007458078,
+      "learning_rate": 9.81848675912457e-06,
+      "loss": 0.5473,
+      "step": 1781
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5649728998215756,
+      "learning_rate": 9.818211226117517e-06,
+      "loss": 0.5303,
+      "step": 1782
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7576583466894525,
+      "learning_rate": 9.81793548801461e-06,
+      "loss": 0.6118,
+      "step": 1783
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6930723487360514,
+      "learning_rate": 9.817659544827582e-06,
+      "loss": 0.547,
+      "step": 1784
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8219685663116344,
+      "learning_rate": 9.817383396568185e-06,
+      "loss": 0.6618,
+      "step": 1785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7419186308064855,
+      "learning_rate": 9.817107043248171e-06,
+      "loss": 0.5586,
+      "step": 1786
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6580774179682489,
+      "learning_rate": 9.816830484879302e-06,
+      "loss": 0.604,
+      "step": 1787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6058216236679842,
+      "learning_rate": 9.816553721473352e-06,
+      "loss": 0.566,
+      "step": 1788
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7888254498849562,
+      "learning_rate": 9.816276753042103e-06,
+      "loss": 0.645,
+      "step": 1789
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6141278230424722,
+      "learning_rate": 9.815999579597342e-06,
+      "loss": 0.5402,
+      "step": 1790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6808504418226013,
+      "learning_rate": 9.815722201150869e-06,
+      "loss": 0.5652,
+      "step": 1791
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8140413787945604,
+      "learning_rate": 9.815444617714489e-06,
+      "loss": 0.6463,
+      "step": 1792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5419997332132857,
+      "learning_rate": 9.815166829300022e-06,
+      "loss": 0.4861,
+      "step": 1793
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6864311376367265,
+      "learning_rate": 9.81488883591929e-06,
+      "loss": 0.5347,
+      "step": 1794
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7049230119572016,
+      "learning_rate": 9.814610637584125e-06,
+      "loss": 0.5836,
+      "step": 1795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6704913119275552,
+      "learning_rate": 9.81433223430637e-06,
+      "loss": 0.5475,
+      "step": 1796
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6143483032874669,
+      "learning_rate": 9.814053626097879e-06,
+      "loss": 0.5557,
+      "step": 1797
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6345160268573378,
+      "learning_rate": 9.813774812970506e-06,
+      "loss": 0.5294,
+      "step": 1798
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8201777018395332,
+      "learning_rate": 9.813495794936123e-06,
+      "loss": 0.6206,
+      "step": 1799
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0710772418338836,
+      "learning_rate": 9.813216572006606e-06,
+      "loss": 0.664,
+      "step": 1800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7240729327012577,
+      "learning_rate": 9.812937144193839e-06,
+      "loss": 0.6213,
+      "step": 1801
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5999673694543911,
+      "learning_rate": 9.81265751150972e-06,
+      "loss": 0.5311,
+      "step": 1802
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9403646063554219,
+      "learning_rate": 9.812377673966147e-06,
+      "loss": 0.666,
+      "step": 1803
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6040105186180103,
+      "learning_rate": 9.812097631575037e-06,
+      "loss": 0.5258,
+      "step": 1804
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0974913112250428,
+      "learning_rate": 9.811817384348306e-06,
+      "loss": 0.6727,
+      "step": 1805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8672915396636229,
+      "learning_rate": 9.811536932297887e-06,
+      "loss": 0.5533,
+      "step": 1806
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6340230627965114,
+      "learning_rate": 9.811256275435716e-06,
+      "loss": 0.5834,
+      "step": 1807
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8718616880804044,
+      "learning_rate": 9.81097541377374e-06,
+      "loss": 0.6046,
+      "step": 1808
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7195864741878601,
+      "learning_rate": 9.810694347323913e-06,
+      "loss": 0.5896,
+      "step": 1809
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.834375755837981,
+      "learning_rate": 9.810413076098201e-06,
+      "loss": 0.6209,
+      "step": 1810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0459165288147858,
+      "learning_rate": 9.810131600108577e-06,
+      "loss": 0.7057,
+      "step": 1811
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9085264408735688,
+      "learning_rate": 9.809849919367021e-06,
+      "loss": 0.6524,
+      "step": 1812
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6245803370399223,
+      "learning_rate": 9.809568033885523e-06,
+      "loss": 0.5362,
+      "step": 1813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.62992976780188,
+      "learning_rate": 9.809285943676086e-06,
+      "loss": 0.5811,
+      "step": 1814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7588014992492863,
+      "learning_rate": 9.809003648750712e-06,
+      "loss": 0.6019,
+      "step": 1815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8454433772131614,
+      "learning_rate": 9.808721149121421e-06,
+      "loss": 0.6199,
+      "step": 1816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6311787764848752,
+      "learning_rate": 9.808438444800238e-06,
+      "loss": 0.5422,
+      "step": 1817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7647311379504511,
+      "learning_rate": 9.808155535799196e-06,
+      "loss": 0.6039,
+      "step": 1818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6859108418964911,
+      "learning_rate": 9.807872422130336e-06,
+      "loss": 0.5781,
+      "step": 1819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7272285110153806,
+      "learning_rate": 9.80758910380571e-06,
+      "loss": 0.5255,
+      "step": 1820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9282585204580677,
+      "learning_rate": 9.80730558083738e-06,
+      "loss": 0.6676,
+      "step": 1821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7771576518488775,
+      "learning_rate": 9.807021853237415e-06,
+      "loss": 0.6327,
+      "step": 1822
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8991058364926697,
+      "learning_rate": 9.806737921017889e-06,
+      "loss": 0.5972,
+      "step": 1823
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6311085357228988,
+      "learning_rate": 9.80645378419089e-06,
+      "loss": 0.5644,
+      "step": 1824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7044051154406225,
+      "learning_rate": 9.806169442768512e-06,
+      "loss": 0.6508,
+      "step": 1825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.377727237075721,
+      "learning_rate": 9.805884896762861e-06,
+      "loss": 0.6146,
+      "step": 1826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6541557095893447,
+      "learning_rate": 9.805600146186045e-06,
+      "loss": 0.535,
+      "step": 1827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7937694622750345,
+      "learning_rate": 9.805315191050189e-06,
+      "loss": 0.6149,
+      "step": 1828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6882330971401761,
+      "learning_rate": 9.805030031367421e-06,
+      "loss": 0.5604,
+      "step": 1829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5556525189442758,
+      "learning_rate": 9.804744667149878e-06,
+      "loss": 0.4953,
+      "step": 1830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6249957602193592,
+      "learning_rate": 9.80445909840971e-06,
+      "loss": 0.5598,
+      "step": 1831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6452773557447997,
+      "learning_rate": 9.80417332515907e-06,
+      "loss": 0.5606,
+      "step": 1832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.881344799149902,
+      "learning_rate": 9.803887347410123e-06,
+      "loss": 0.617,
+      "step": 1833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7228633029360589,
+      "learning_rate": 9.803601165175043e-06,
+      "loss": 0.5779,
+      "step": 1834
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6328450543531903,
+      "learning_rate": 9.803314778466011e-06,
+      "loss": 0.5072,
+      "step": 1835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6878082965265847,
+      "learning_rate": 9.803028187295218e-06,
+      "loss": 0.5446,
+      "step": 1836
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7561265841790251,
+      "learning_rate": 9.802741391674864e-06,
+      "loss": 0.6078,
+      "step": 1837
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5269171983317034,
+      "learning_rate": 9.802454391617158e-06,
+      "loss": 0.6168,
+      "step": 1838
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.662341341988264,
+      "learning_rate": 9.802167187134312e-06,
+      "loss": 0.5304,
+      "step": 1839
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6361927091599409,
+      "learning_rate": 9.801879778238557e-06,
+      "loss": 0.5468,
+      "step": 1840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8665692683559979,
+      "learning_rate": 9.801592164942122e-06,
+      "loss": 0.654,
+      "step": 1841
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5794446386427597,
+      "learning_rate": 9.801304347257255e-06,
+      "loss": 0.514,
+      "step": 1842
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.565083604405036,
+      "learning_rate": 9.801016325196201e-06,
+      "loss": 0.4536,
+      "step": 1843
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7119210720305069,
+      "learning_rate": 9.800728098771227e-06,
+      "loss": 0.5877,
+      "step": 1844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6911282038193561,
+      "learning_rate": 9.800439667994597e-06,
+      "loss": 0.5466,
+      "step": 1845
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1006291115988809,
+      "learning_rate": 9.800151032878593e-06,
+      "loss": 0.6649,
+      "step": 1846
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7941378228483236,
+      "learning_rate": 9.799862193435496e-06,
+      "loss": 0.6036,
+      "step": 1847
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7007455534210305,
+      "learning_rate": 9.799573149677604e-06,
+      "loss": 0.5798,
+      "step": 1848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7955570218586099,
+      "learning_rate": 9.79928390161722e-06,
+      "loss": 0.6842,
+      "step": 1849
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6485587476088552,
+      "learning_rate": 9.798994449266657e-06,
+      "loss": 0.5086,
+      "step": 1850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7990949334892158,
+      "learning_rate": 9.798704792638235e-06,
+      "loss": 0.6534,
+      "step": 1851
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6409614614806528,
+      "learning_rate": 9.798414931744286e-06,
+      "loss": 0.5419,
+      "step": 1852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8167959177032953,
+      "learning_rate": 9.798124866597148e-06,
+      "loss": 0.6005,
+      "step": 1853
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.653313874850728,
+      "learning_rate": 9.797834597209164e-06,
+      "loss": 0.5686,
+      "step": 1854
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7840320310347608,
+      "learning_rate": 9.797544123592694e-06,
+      "loss": 0.6048,
+      "step": 1855
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6548722236822639,
+      "learning_rate": 9.797253445760103e-06,
+      "loss": 0.471,
+      "step": 1856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8512904379792536,
+      "learning_rate": 9.796962563723761e-06,
+      "loss": 0.6613,
+      "step": 1857
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9699865123154551,
+      "learning_rate": 9.796671477496052e-06,
+      "loss": 0.7084,
+      "step": 1858
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2359218981242521,
+      "learning_rate": 9.796380187089365e-06,
+      "loss": 0.7023,
+      "step": 1859
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6176424523894019,
+      "learning_rate": 9.7960886925161e-06,
+      "loss": 0.5307,
+      "step": 1860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6457671000241276,
+      "learning_rate": 9.795796993788669e-06,
+      "loss": 0.5582,
+      "step": 1861
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8170632193767384,
+      "learning_rate": 9.795505090919483e-06,
+      "loss": 0.6949,
+      "step": 1862
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6247440359342894,
+      "learning_rate": 9.79521298392097e-06,
+      "loss": 0.5519,
+      "step": 1863
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8991623154905343,
+      "learning_rate": 9.794920672805563e-06,
+      "loss": 0.6027,
+      "step": 1864
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6371668104566895,
+      "learning_rate": 9.794628157585705e-06,
+      "loss": 0.5482,
+      "step": 1865
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7367620987083294,
+      "learning_rate": 9.794335438273847e-06,
+      "loss": 0.6282,
+      "step": 1866
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8788005777652736,
+      "learning_rate": 9.794042514882453e-06,
+      "loss": 0.6181,
+      "step": 1867
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2160155982655554,
+      "learning_rate": 9.793749387423986e-06,
+      "loss": 0.6313,
+      "step": 1868
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6875203175738123,
+      "learning_rate": 9.793456055910929e-06,
+      "loss": 0.5657,
+      "step": 1869
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.620390393773217,
+      "learning_rate": 9.79316252035576e-06,
+      "loss": 0.5636,
+      "step": 1870
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8060927637625144,
+      "learning_rate": 9.792868780770984e-06,
+      "loss": 0.6076,
+      "step": 1871
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7892093311624908,
+      "learning_rate": 9.7925748371691e-06,
+      "loss": 0.6325,
+      "step": 1872
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6340907113812765,
+      "learning_rate": 9.792280689562617e-06,
+      "loss": 0.5467,
+      "step": 1873
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7812304518114844,
+      "learning_rate": 9.791986337964061e-06,
+      "loss": 0.5487,
+      "step": 1874
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6868687819751733,
+      "learning_rate": 9.79169178238596e-06,
+      "loss": 0.5666,
+      "step": 1875
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7470990444460175,
+      "learning_rate": 9.79139702284085e-06,
+      "loss": 0.5775,
+      "step": 1876
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.669118934532083,
+      "learning_rate": 9.791102059341282e-06,
+      "loss": 0.5375,
+      "step": 1877
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5789565813247094,
+      "learning_rate": 9.790806891899807e-06,
+      "loss": 0.5067,
+      "step": 1878
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8735880122952258,
+      "learning_rate": 9.790511520528994e-06,
+      "loss": 0.655,
+      "step": 1879
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7434939390969325,
+      "learning_rate": 9.790215945241414e-06,
+      "loss": 0.6341,
+      "step": 1880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6660027918812905,
+      "learning_rate": 9.789920166049646e-06,
+      "loss": 0.5673,
+      "step": 1881
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8022913933066801,
+      "learning_rate": 9.789624182966287e-06,
+      "loss": 0.6094,
+      "step": 1882
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6770294527721105,
+      "learning_rate": 9.789327996003928e-06,
+      "loss": 0.5774,
+      "step": 1883
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.94355827439604,
+      "learning_rate": 9.789031605175183e-06,
+      "loss": 0.6976,
+      "step": 1884
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7256597176856311,
+      "learning_rate": 9.788735010492665e-06,
+      "loss": 0.5596,
+      "step": 1885
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8615465433468593,
+      "learning_rate": 9.788438211969e-06,
+      "loss": 0.6355,
+      "step": 1886
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8250046633773639,
+      "learning_rate": 9.788141209616823e-06,
+      "loss": 0.6403,
+      "step": 1887
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7262155352396545,
+      "learning_rate": 9.787844003448776e-06,
+      "loss": 0.6173,
+      "step": 1888
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6120776219419543,
+      "learning_rate": 9.787546593477509e-06,
+      "loss": 0.555,
+      "step": 1889
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5843370001787598,
+      "learning_rate": 9.78724897971568e-06,
+      "loss": 0.5397,
+      "step": 1890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7438143537722247,
+      "learning_rate": 9.786951162175961e-06,
+      "loss": 0.6256,
+      "step": 1891
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6842285821982648,
+      "learning_rate": 9.786653140871029e-06,
+      "loss": 0.5712,
+      "step": 1892
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.080874953127134,
+      "learning_rate": 9.78635491581357e-06,
+      "loss": 0.6835,
+      "step": 1893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7851095478476103,
+      "learning_rate": 9.786056487016276e-06,
+      "loss": 0.5984,
+      "step": 1894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9732872864863693,
+      "learning_rate": 9.785757854491851e-06,
+      "loss": 0.6524,
+      "step": 1895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6817606329799621,
+      "learning_rate": 9.785459018253007e-06,
+      "loss": 0.5323,
+      "step": 1896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5777681519472451,
+      "learning_rate": 9.785159978312465e-06,
+      "loss": 0.5827,
+      "step": 1897
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9055026562937675,
+      "learning_rate": 9.784860734682954e-06,
+      "loss": 0.7013,
+      "step": 1898
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6536385790471594,
+      "learning_rate": 9.78456128737721e-06,
+      "loss": 0.5104,
+      "step": 1899
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6870498844950168,
+      "learning_rate": 9.784261636407983e-06,
+      "loss": 0.5769,
+      "step": 1900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.564708476266616,
+      "learning_rate": 9.783961781788027e-06,
+      "loss": 0.4899,
+      "step": 1901
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6478259646680682,
+      "learning_rate": 9.783661723530105e-06,
+      "loss": 0.5143,
+      "step": 1902
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5663585499776949,
+      "learning_rate": 9.78336146164699e-06,
+      "loss": 0.5137,
+      "step": 1903
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5878041448139183,
+      "learning_rate": 9.783060996151461e-06,
+      "loss": 0.5116,
+      "step": 1904
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7619117929481182,
+      "learning_rate": 9.78276032705631e-06,
+      "loss": 0.6297,
+      "step": 1905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8115502761321896,
+      "learning_rate": 9.782459454374337e-06,
+      "loss": 0.5855,
+      "step": 1906
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7862715666596223,
+      "learning_rate": 9.78215837811835e-06,
+      "loss": 0.6286,
+      "step": 1907
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.681053413459562,
+      "learning_rate": 9.781857098301158e-06,
+      "loss": 0.5786,
+      "step": 1908
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5921885004324525,
+      "learning_rate": 9.781555614935591e-06,
+      "loss": 0.5684,
+      "step": 1909
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.001601564863681,
+      "learning_rate": 9.781253928034483e-06,
+      "loss": 0.6666,
+      "step": 1910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5639342154583687,
+      "learning_rate": 9.780952037610674e-06,
+      "loss": 0.4926,
+      "step": 1911
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.948207088170478,
+      "learning_rate": 9.780649943677013e-06,
+      "loss": 0.6486,
+      "step": 1912
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.731045565932882,
+      "learning_rate": 9.780347646246359e-06,
+      "loss": 0.6269,
+      "step": 1913
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7373371843437875,
+      "learning_rate": 9.780045145331585e-06,
+      "loss": 0.5573,
+      "step": 1914
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8560864430134707,
+      "learning_rate": 9.779742440945563e-06,
+      "loss": 0.64,
+      "step": 1915
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3035260561724835,
+      "learning_rate": 9.779439533101178e-06,
+      "loss": 0.6405,
+      "step": 1916
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8829823219148513,
+      "learning_rate": 9.779136421811326e-06,
+      "loss": 0.6577,
+      "step": 1917
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9189240436124142,
+      "learning_rate": 9.778833107088907e-06,
+      "loss": 0.6538,
+      "step": 1918
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8070312100696989,
+      "learning_rate": 9.778529588946835e-06,
+      "loss": 0.6585,
+      "step": 1919
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6162431523006736,
+      "learning_rate": 9.778225867398027e-06,
+      "loss": 0.5574,
+      "step": 1920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7038298018322586,
+      "learning_rate": 9.777921942455414e-06,
+      "loss": 0.5793,
+      "step": 1921
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8130947236153864,
+      "learning_rate": 9.777617814131934e-06,
+      "loss": 0.6469,
+      "step": 1922
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6830042517603784,
+      "learning_rate": 9.777313482440528e-06,
+      "loss": 0.5905,
+      "step": 1923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7720600364389372,
+      "learning_rate": 9.777008947394152e-06,
+      "loss": 0.625,
+      "step": 1924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7757763888398167,
+      "learning_rate": 9.776704209005774e-06,
+      "loss": 0.6185,
+      "step": 1925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7150199268120511,
+      "learning_rate": 9.77639926728836e-06,
+      "loss": 0.5528,
+      "step": 1926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6747235250631582,
+      "learning_rate": 9.776094122254892e-06,
+      "loss": 0.6118,
+      "step": 1927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5501730011096506,
+      "learning_rate": 9.77578877391836e-06,
+      "loss": 0.5672,
+      "step": 1928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6747575054042125,
+      "learning_rate": 9.775483222291762e-06,
+      "loss": 0.579,
+      "step": 1929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6437417447268526,
+      "learning_rate": 9.775177467388101e-06,
+      "loss": 0.5662,
+      "step": 1930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5818262316647661,
+      "learning_rate": 9.774871509220396e-06,
+      "loss": 0.5539,
+      "step": 1931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6384674367901276,
+      "learning_rate": 9.77456534780167e-06,
+      "loss": 0.4779,
+      "step": 1932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6626283438878735,
+      "learning_rate": 9.774258983144952e-06,
+      "loss": 0.5772,
+      "step": 1933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0488726515835534,
+      "learning_rate": 9.773952415263288e-06,
+      "loss": 0.6297,
+      "step": 1934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6240739938305208,
+      "learning_rate": 9.773645644169724e-06,
+      "loss": 0.5175,
+      "step": 1935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8390389841799185,
+      "learning_rate": 9.77333866987732e-06,
+      "loss": 0.6576,
+      "step": 1936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8315159132638323,
+      "learning_rate": 9.77303149239914e-06,
+      "loss": 0.5954,
+      "step": 1937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6388266061400573,
+      "learning_rate": 9.772724111748265e-06,
+      "loss": 0.5862,
+      "step": 1938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1806161302822178,
+      "learning_rate": 9.772416527937774e-06,
+      "loss": 0.6431,
+      "step": 1939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7005717575610654,
+      "learning_rate": 9.772108740980764e-06,
+      "loss": 0.5804,
+      "step": 1940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5913186514354882,
+      "learning_rate": 9.77180075089033e-06,
+      "loss": 0.5499,
+      "step": 1941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6922462934160454,
+      "learning_rate": 9.77149255767959e-06,
+      "loss": 0.5241,
+      "step": 1942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.77347767174598,
+      "learning_rate": 9.77118416136166e-06,
+      "loss": 0.6407,
+      "step": 1943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8038559729479153,
+      "learning_rate": 9.770875561949665e-06,
+      "loss": 0.6179,
+      "step": 1944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6279371308770109,
+      "learning_rate": 9.770566759456743e-06,
+      "loss": 0.5585,
+      "step": 1945
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7289729212128583,
+      "learning_rate": 9.770257753896038e-06,
+      "loss": 0.5612,
+      "step": 1946
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.615149968987624,
+      "learning_rate": 9.769948545280705e-06,
+      "loss": 0.5887,
+      "step": 1947
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6851751966271976,
+      "learning_rate": 9.769639133623905e-06,
+      "loss": 0.5275,
+      "step": 1948
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5536769005121086,
+      "learning_rate": 9.769329518938808e-06,
+      "loss": 0.5175,
+      "step": 1949
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6095844296252171,
+      "learning_rate": 9.769019701238595e-06,
+      "loss": 0.508,
+      "step": 1950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6736914616040667,
+      "learning_rate": 9.768709680536453e-06,
+      "loss": 0.5366,
+      "step": 1951
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.637199326869723,
+      "learning_rate": 9.768399456845577e-06,
+      "loss": 0.5953,
+      "step": 1952
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8049306836349139,
+      "learning_rate": 9.768089030179175e-06,
+      "loss": 0.583,
+      "step": 1953
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6303732958697623,
+      "learning_rate": 9.767778400550459e-06,
+      "loss": 0.5499,
+      "step": 1954
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6099271068664792,
+      "learning_rate": 9.767467567972653e-06,
+      "loss": 0.59,
+      "step": 1955
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7435821689088031,
+      "learning_rate": 9.767156532458985e-06,
+      "loss": 0.6661,
+      "step": 1956
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7438129889723072,
+      "learning_rate": 9.7668452940227e-06,
+      "loss": 0.6034,
+      "step": 1957
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.112723556132667,
+      "learning_rate": 9.766533852677042e-06,
+      "loss": 0.5991,
+      "step": 1958
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6311477087148216,
+      "learning_rate": 9.76622220843527e-06,
+      "loss": 0.5811,
+      "step": 1959
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6599377713956334,
+      "learning_rate": 9.765910361310648e-06,
+      "loss": 0.5685,
+      "step": 1960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6476688505922261,
+      "learning_rate": 9.765598311316453e-06,
+      "loss": 0.5605,
+      "step": 1961
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0305340753836008,
+      "learning_rate": 9.765286058465966e-06,
+      "loss": 0.6607,
+      "step": 1962
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.044994744196121,
+      "learning_rate": 9.764973602772479e-06,
+      "loss": 0.6493,
+      "step": 1963
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6759009554840261,
+      "learning_rate": 9.764660944249294e-06,
+      "loss": 0.5395,
+      "step": 1964
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6588819199833671,
+      "learning_rate": 9.764348082909716e-06,
+      "loss": 0.5729,
+      "step": 1965
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5676185798033166,
+      "learning_rate": 9.764035018767066e-06,
+      "loss": 0.6366,
+      "step": 1966
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7488411094021024,
+      "learning_rate": 9.763721751834669e-06,
+      "loss": 0.6195,
+      "step": 1967
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7798042275961441,
+      "learning_rate": 9.763408282125859e-06,
+      "loss": 0.6319,
+      "step": 1968
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6113906584451468,
+      "learning_rate": 9.76309460965398e-06,
+      "loss": 0.5173,
+      "step": 1969
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9364621296730903,
+      "learning_rate": 9.762780734432385e-06,
+      "loss": 0.6356,
+      "step": 1970
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5893345078880406,
+      "learning_rate": 9.762466656474434e-06,
+      "loss": 0.5172,
+      "step": 1971
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9155751022027765,
+      "learning_rate": 9.762152375793494e-06,
+      "loss": 0.5782,
+      "step": 1972
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5915185565742986,
+      "learning_rate": 9.761837892402947e-06,
+      "loss": 0.4789,
+      "step": 1973
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7598180720449295,
+      "learning_rate": 9.761523206316178e-06,
+      "loss": 0.6276,
+      "step": 1974
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6946721467354218,
+      "learning_rate": 9.76120831754658e-06,
+      "loss": 0.5663,
+      "step": 1975
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7861411569330273,
+      "learning_rate": 9.76089322610756e-06,
+      "loss": 0.6307,
+      "step": 1976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7665265029916419,
+      "learning_rate": 9.760577932012529e-06,
+      "loss": 0.6186,
+      "step": 1977
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6485763026996052,
+      "learning_rate": 9.760262435274907e-06,
+      "loss": 0.5554,
+      "step": 1978
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0154239397820972,
+      "learning_rate": 9.759946735908125e-06,
+      "loss": 0.6091,
+      "step": 1979
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6333491185463955,
+      "learning_rate": 9.759630833925622e-06,
+      "loss": 0.5214,
+      "step": 1980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8024634694331682,
+      "learning_rate": 9.759314729340843e-06,
+      "loss": 0.5984,
+      "step": 1981
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5994701254065763,
+      "learning_rate": 9.758998422167245e-06,
+      "loss": 0.5424,
+      "step": 1982
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8043182819282072,
+      "learning_rate": 9.758681912418292e-06,
+      "loss": 0.6087,
+      "step": 1983
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6638797433462842,
+      "learning_rate": 9.758365200107455e-06,
+      "loss": 0.5653,
+      "step": 1984
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8512709591686907,
+      "learning_rate": 9.75804828524822e-06,
+      "loss": 0.6407,
+      "step": 1985
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7922522876482098,
+      "learning_rate": 9.757731167854072e-06,
+      "loss": 0.5331,
+      "step": 1986
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6068064293105282,
+      "learning_rate": 9.757413847938512e-06,
+      "loss": 0.5123,
+      "step": 1987
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8246795405287417,
+      "learning_rate": 9.757096325515047e-06,
+      "loss": 0.6201,
+      "step": 1988
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9882681574025354,
+      "learning_rate": 9.756778600597193e-06,
+      "loss": 0.6364,
+      "step": 1989
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0554044969170837,
+      "learning_rate": 9.756460673198474e-06,
+      "loss": 0.587,
+      "step": 1990
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8535287115138298,
+      "learning_rate": 9.756142543332424e-06,
+      "loss": 0.6933,
+      "step": 1991
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7035383973149673,
+      "learning_rate": 9.755824211012585e-06,
+      "loss": 0.5658,
+      "step": 1992
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6204193016440039,
+      "learning_rate": 9.755505676252506e-06,
+      "loss": 0.497,
+      "step": 1993
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.757533191159159,
+      "learning_rate": 9.755186939065746e-06,
+      "loss": 0.6107,
+      "step": 1994
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6809709597015329,
+      "learning_rate": 9.754867999465876e-06,
+      "loss": 0.5836,
+      "step": 1995
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6514669512667115,
+      "learning_rate": 9.754548857466468e-06,
+      "loss": 0.549,
+      "step": 1996
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5753757412548579,
+      "learning_rate": 9.754229513081109e-06,
+      "loss": 0.5024,
+      "step": 1997
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.898139757103797,
+      "learning_rate": 9.753909966323389e-06,
+      "loss": 0.6627,
+      "step": 1998
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7743361101681899,
+      "learning_rate": 9.753590217206917e-06,
+      "loss": 0.5513,
+      "step": 1999
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8926775684335282,
+      "learning_rate": 9.753270265745298e-06,
+      "loss": 0.6801,
+      "step": 2000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.905779177290222,
+      "learning_rate": 9.752950111952153e-06,
+      "loss": 0.6537,
+      "step": 2001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.587992147442533,
+      "learning_rate": 9.75262975584111e-06,
+      "loss": 0.4905,
+      "step": 2002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6776931912155311,
+      "learning_rate": 9.752309197425807e-06,
+      "loss": 0.5584,
+      "step": 2003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9758447500493901,
+      "learning_rate": 9.751988436719886e-06,
+      "loss": 0.6163,
+      "step": 2004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.786220018553936,
+      "learning_rate": 9.751667473737003e-06,
+      "loss": 0.6614,
+      "step": 2005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8054724299259647,
+      "learning_rate": 9.751346308490819e-06,
+      "loss": 0.6208,
+      "step": 2006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6953486716148269,
+      "learning_rate": 9.751024940995008e-06,
+      "loss": 0.5441,
+      "step": 2007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7116303093210296,
+      "learning_rate": 9.750703371263246e-06,
+      "loss": 0.5668,
+      "step": 2008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6128615339599108,
+      "learning_rate": 9.750381599309223e-06,
+      "loss": 0.5103,
+      "step": 2009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5895524545957406,
+      "learning_rate": 9.750059625146634e-06,
+      "loss": 0.5451,
+      "step": 2010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7106416463269892,
+      "learning_rate": 9.749737448789188e-06,
+      "loss": 0.5748,
+      "step": 2011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9539169606527462,
+      "learning_rate": 9.749415070250595e-06,
+      "loss": 0.616,
+      "step": 2012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8121783756099556,
+      "learning_rate": 9.749092489544578e-06,
+      "loss": 0.661,
+      "step": 2013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5523981654354464,
+      "learning_rate": 9.748769706684872e-06,
+      "loss": 0.4868,
+      "step": 2014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9279060732475999,
+      "learning_rate": 9.748446721685214e-06,
+      "loss": 0.67,
+      "step": 2015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6502826087472836,
+      "learning_rate": 9.748123534559353e-06,
+      "loss": 0.5581,
+      "step": 2016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9561047719262271,
+      "learning_rate": 9.747800145321047e-06,
+      "loss": 0.6016,
+      "step": 2017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6914834687803977,
+      "learning_rate": 9.74747655398406e-06,
+      "loss": 0.5959,
+      "step": 2018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9461309182534641,
+      "learning_rate": 9.747152760562165e-06,
+      "loss": 0.6504,
+      "step": 2019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6677868296472617,
+      "learning_rate": 9.746828765069148e-06,
+      "loss": 0.5674,
+      "step": 2020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7205211983597987,
+      "learning_rate": 9.7465045675188e-06,
+      "loss": 0.5707,
+      "step": 2021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8565289265491989,
+      "learning_rate": 9.746180167924919e-06,
+      "loss": 0.5994,
+      "step": 2022
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7827491188198623,
+      "learning_rate": 9.745855566301315e-06,
+      "loss": 0.6366,
+      "step": 2023
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8513641718980255,
+      "learning_rate": 9.745530762661805e-06,
+      "loss": 0.6954,
+      "step": 2024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9325634003810489,
+      "learning_rate": 9.745205757020216e-06,
+      "loss": 0.6333,
+      "step": 2025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6288848026923852,
+      "learning_rate": 9.744880549390382e-06,
+      "loss": 0.5053,
+      "step": 2026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7933685981261801,
+      "learning_rate": 9.744555139786142e-06,
+      "loss": 0.5391,
+      "step": 2027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8354595558937188,
+      "learning_rate": 9.744229528221354e-06,
+      "loss": 0.572,
+      "step": 2028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6365256178368215,
+      "learning_rate": 9.743903714709875e-06,
+      "loss": 0.5329,
+      "step": 2029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0296411852335627,
+      "learning_rate": 9.743577699265574e-06,
+      "loss": 0.694,
+      "step": 2030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5709720494840266,
+      "learning_rate": 9.743251481902329e-06,
+      "loss": 0.5204,
+      "step": 2031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7712303990324043,
+      "learning_rate": 9.742925062634025e-06,
+      "loss": 0.5535,
+      "step": 2032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.036540970751146,
+      "learning_rate": 9.742598441474558e-06,
+      "loss": 0.6278,
+      "step": 2033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7719481581583385,
+      "learning_rate": 9.74227161843783e-06,
+      "loss": 0.6635,
+      "step": 2034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7193126175204697,
+      "learning_rate": 9.741944593537754e-06,
+      "loss": 0.5871,
+      "step": 2035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5914133858909755,
+      "learning_rate": 9.74161736678825e-06,
+      "loss": 0.5482,
+      "step": 2036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8380293249144736,
+      "learning_rate": 9.741289938203246e-06,
+      "loss": 0.5469,
+      "step": 2037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.717292077984541,
+      "learning_rate": 9.74096230779668e-06,
+      "loss": 0.5984,
+      "step": 2038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8120503871855917,
+      "learning_rate": 9.740634475582499e-06,
+      "loss": 0.6502,
+      "step": 2039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6498869588587249,
+      "learning_rate": 9.740306441574656e-06,
+      "loss": 0.5655,
+      "step": 2040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.650287790228828,
+      "learning_rate": 9.739978205787117e-06,
+      "loss": 0.5129,
+      "step": 2041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4853959479523764,
+      "learning_rate": 9.739649768233853e-06,
+      "loss": 0.5819,
+      "step": 2042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6446973900005055,
+      "learning_rate": 9.739321128928844e-06,
+      "loss": 0.5536,
+      "step": 2043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6946298845745377,
+      "learning_rate": 9.738992287886078e-06,
+      "loss": 0.5143,
+      "step": 2044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5753733402622996,
+      "learning_rate": 9.738663245119555e-06,
+      "loss": 0.5252,
+      "step": 2045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7782048941045351,
+      "learning_rate": 9.738334000643278e-06,
+      "loss": 0.5927,
+      "step": 2046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9996804655740105,
+      "learning_rate": 9.738004554471267e-06,
+      "loss": 0.6152,
+      "step": 2047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8691303856697805,
+      "learning_rate": 9.737674906617543e-06,
+      "loss": 0.6083,
+      "step": 2048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6258160618603322,
+      "learning_rate": 9.737345057096135e-06,
+      "loss": 0.5168,
+      "step": 2049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.775293773299195,
+      "learning_rate": 9.737015005921088e-06,
+      "loss": 0.6381,
+      "step": 2050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6385407024821724,
+      "learning_rate": 9.73668475310645e-06,
+      "loss": 0.5543,
+      "step": 2051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5861410800895339,
+      "learning_rate": 9.736354298666277e-06,
+      "loss": 0.4962,
+      "step": 2052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9199703456080394,
+      "learning_rate": 9.736023642614638e-06,
+      "loss": 0.6674,
+      "step": 2053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9154313901885083,
+      "learning_rate": 9.735692784965606e-06,
+      "loss": 0.6767,
+      "step": 2054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1384153897500089,
+      "learning_rate": 9.735361725733265e-06,
+      "loss": 0.6438,
+      "step": 2055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6542157276954875,
+      "learning_rate": 9.735030464931707e-06,
+      "loss": 0.5181,
+      "step": 2056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7714882273538483,
+      "learning_rate": 9.734699002575035e-06,
+      "loss": 0.6218,
+      "step": 2057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9906874791748633,
+      "learning_rate": 9.734367338677355e-06,
+      "loss": 0.6813,
+      "step": 2058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6833249408232183,
+      "learning_rate": 9.734035473252786e-06,
+      "loss": 0.5471,
+      "step": 2059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6003145554045849,
+      "learning_rate": 9.733703406315455e-06,
+      "loss": 0.5351,
+      "step": 2060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7307122835592621,
+      "learning_rate": 9.733371137879498e-06,
+      "loss": 0.5821,
+      "step": 2061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6831646478598287,
+      "learning_rate": 9.733038667959054e-06,
+      "loss": 0.5798,
+      "step": 2062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5258412523513408,
+      "learning_rate": 9.73270599656828e-06,
+      "loss": 0.4822,
+      "step": 2063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.288210033872429,
+      "learning_rate": 9.732373123721337e-06,
+      "loss": 0.635,
+      "step": 2064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6161080538188768,
+      "learning_rate": 9.732040049432393e-06,
+      "loss": 0.5423,
+      "step": 2065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9489274096999086,
+      "learning_rate": 9.731706773715624e-06,
+      "loss": 0.6287,
+      "step": 2066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6266522673051259,
+      "learning_rate": 9.731373296585218e-06,
+      "loss": 0.5672,
+      "step": 2067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6733008292071735,
+      "learning_rate": 9.73103961805537e-06,
+      "loss": 0.6035,
+      "step": 2068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.722691724097056,
+      "learning_rate": 9.730705738140284e-06,
+      "loss": 0.6052,
+      "step": 2069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5733938543934274,
+      "learning_rate": 9.730371656854172e-06,
+      "loss": 0.4776,
+      "step": 2070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.620428591961551,
+      "learning_rate": 9.730037374211255e-06,
+      "loss": 0.5432,
+      "step": 2071
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7010401930330775,
+      "learning_rate": 9.729702890225761e-06,
+      "loss": 0.5135,
+      "step": 2072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6290161986027752,
+      "learning_rate": 9.729368204911928e-06,
+      "loss": 0.5256,
+      "step": 2073
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6798955498123993,
+      "learning_rate": 9.729033318284005e-06,
+      "loss": 0.5764,
+      "step": 2074
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7569271787854982,
+      "learning_rate": 9.728698230356246e-06,
+      "loss": 0.5505,
+      "step": 2075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6169296838439209,
+      "learning_rate": 9.728362941142913e-06,
+      "loss": 0.5342,
+      "step": 2076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.665497242148675,
+      "learning_rate": 9.72802745065828e-06,
+      "loss": 0.559,
+      "step": 2077
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8047324862964884,
+      "learning_rate": 9.727691758916627e-06,
+      "loss": 0.6709,
+      "step": 2078
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6871136990546536,
+      "learning_rate": 9.727355865932242e-06,
+      "loss": 0.5583,
+      "step": 2079
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6454370635444046,
+      "learning_rate": 9.727019771719427e-06,
+      "loss": 0.522,
+      "step": 2080
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6001209377809871,
+      "learning_rate": 9.726683476292484e-06,
+      "loss": 0.542,
+      "step": 2081
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6933759361085206,
+      "learning_rate": 9.72634697966573e-06,
+      "loss": 0.5507,
+      "step": 2082
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9055509903938554,
+      "learning_rate": 9.726010281853488e-06,
+      "loss": 0.6022,
+      "step": 2083
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6116977883299418,
+      "learning_rate": 9.725673382870092e-06,
+      "loss": 0.5139,
+      "step": 2084
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6370131943565843,
+      "learning_rate": 9.725336282729877e-06,
+      "loss": 0.5515,
+      "step": 2085
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9004152023727671,
+      "learning_rate": 9.7249989814472e-06,
+      "loss": 0.6576,
+      "step": 2086
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0158407119320836,
+      "learning_rate": 9.724661479036414e-06,
+      "loss": 0.5774,
+      "step": 2087
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7675904043554551,
+      "learning_rate": 9.724323775511888e-06,
+      "loss": 0.6786,
+      "step": 2088
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7489623594684662,
+      "learning_rate": 9.723985870887995e-06,
+      "loss": 0.5578,
+      "step": 2089
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6084731557722258,
+      "learning_rate": 9.723647765179119e-06,
+      "loss": 0.5833,
+      "step": 2090
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8331052745998051,
+      "learning_rate": 9.723309458399652e-06,
+      "loss": 0.6462,
+      "step": 2091
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7359928436991852,
+      "learning_rate": 9.722970950563995e-06,
+      "loss": 0.5608,
+      "step": 2092
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6837342295014889,
+      "learning_rate": 9.722632241686559e-06,
+      "loss": 0.5407,
+      "step": 2093
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7729308964984413,
+      "learning_rate": 9.722293331781758e-06,
+      "loss": 0.5863,
+      "step": 2094
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8614004223480878,
+      "learning_rate": 9.721954220864019e-06,
+      "loss": 0.6489,
+      "step": 2095
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6232329294269511,
+      "learning_rate": 9.721614908947781e-06,
+      "loss": 0.5391,
+      "step": 2096
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6077719243223584,
+      "learning_rate": 9.721275396047483e-06,
+      "loss": 0.5393,
+      "step": 2097
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.558351867007384,
+      "learning_rate": 9.720935682177577e-06,
+      "loss": 0.4888,
+      "step": 2098
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8062406355126962,
+      "learning_rate": 9.720595767352527e-06,
+      "loss": 0.673,
+      "step": 2099
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9037491752296137,
+      "learning_rate": 9.720255651586799e-06,
+      "loss": 0.6714,
+      "step": 2100
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7360739154254519,
+      "learning_rate": 9.719915334894871e-06,
+      "loss": 0.6018,
+      "step": 2101
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7831402654138032,
+      "learning_rate": 9.71957481729123e-06,
+      "loss": 0.5486,
+      "step": 2102
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0765839399703987,
+      "learning_rate": 9.719234098790374e-06,
+      "loss": 0.5915,
+      "step": 2103
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7966734872967898,
+      "learning_rate": 9.718893179406798e-06,
+      "loss": 0.61,
+      "step": 2104
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.971240792968848,
+      "learning_rate": 9.718552059155022e-06,
+      "loss": 0.6608,
+      "step": 2105
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5970469956752246,
+      "learning_rate": 9.718210738049563e-06,
+      "loss": 0.4595,
+      "step": 2106
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5920103825747084,
+      "learning_rate": 9.71786921610495e-06,
+      "loss": 0.5117,
+      "step": 2107
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8117674785784921,
+      "learning_rate": 9.71752749333572e-06,
+      "loss": 0.6788,
+      "step": 2108
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5256254889132973,
+      "learning_rate": 9.717185569756419e-06,
+      "loss": 0.46,
+      "step": 2109
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5834368545417268,
+      "learning_rate": 9.716843445381603e-06,
+      "loss": 0.5359,
+      "step": 2110
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0017053379236673,
+      "learning_rate": 9.716501120225834e-06,
+      "loss": 0.6183,
+      "step": 2111
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6548945424440169,
+      "learning_rate": 9.716158594303685e-06,
+      "loss": 0.5459,
+      "step": 2112
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.23223843284866,
+      "learning_rate": 9.715815867629735e-06,
+      "loss": 0.6646,
+      "step": 2113
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.659091643792312,
+      "learning_rate": 9.715472940218573e-06,
+      "loss": 0.5551,
+      "step": 2114
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6976834670146201,
+      "learning_rate": 9.715129812084795e-06,
+      "loss": 0.5258,
+      "step": 2115
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8699304617573581,
+      "learning_rate": 9.71478648324301e-06,
+      "loss": 0.6653,
+      "step": 2116
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8140126412568707,
+      "learning_rate": 9.71444295370783e-06,
+      "loss": 0.6282,
+      "step": 2117
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5488027208872968,
+      "learning_rate": 9.71409922349388e-06,
+      "loss": 0.5297,
+      "step": 2118
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7111858818212421,
+      "learning_rate": 9.713755292615789e-06,
+      "loss": 0.5571,
+      "step": 2119
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8020520271094485,
+      "learning_rate": 9.713411161088198e-06,
+      "loss": 0.6039,
+      "step": 2120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7866973296861831,
+      "learning_rate": 9.713066828925757e-06,
+      "loss": 0.6567,
+      "step": 2121
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.642506004797298,
+      "learning_rate": 9.712722296143121e-06,
+      "loss": 0.5614,
+      "step": 2122
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7503891770795013,
+      "learning_rate": 9.712377562754957e-06,
+      "loss": 0.6174,
+      "step": 2123
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7500400246988823,
+      "learning_rate": 9.712032628775939e-06,
+      "loss": 0.5557,
+      "step": 2124
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6542104740649382,
+      "learning_rate": 9.711687494220748e-06,
+      "loss": 0.5911,
+      "step": 2125
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0160392181510352,
+      "learning_rate": 9.711342159104078e-06,
+      "loss": 0.6453,
+      "step": 2126
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7892779950190293,
+      "learning_rate": 9.710996623440627e-06,
+      "loss": 0.5836,
+      "step": 2127
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7363641448666912,
+      "learning_rate": 9.710650887245103e-06,
+      "loss": 0.5829,
+      "step": 2128
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8431743530457255,
+      "learning_rate": 9.710304950532225e-06,
+      "loss": 0.5804,
+      "step": 2129
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8824858658654475,
+      "learning_rate": 9.709958813316718e-06,
+      "loss": 0.6999,
+      "step": 2130
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8998131890206996,
+      "learning_rate": 9.709612475613315e-06,
+      "loss": 0.6894,
+      "step": 2131
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5910197690092174,
+      "learning_rate": 9.709265937436758e-06,
+      "loss": 0.5738,
+      "step": 2132
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.019065391040781,
+      "learning_rate": 9.708919198801799e-06,
+      "loss": 0.6415,
+      "step": 2133
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6614502965096164,
+      "learning_rate": 9.708572259723198e-06,
+      "loss": 0.5642,
+      "step": 2134
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8262117858413601,
+      "learning_rate": 9.70822512021572e-06,
+      "loss": 0.6035,
+      "step": 2135
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9290103171899875,
+      "learning_rate": 9.707877780294147e-06,
+      "loss": 0.6671,
+      "step": 2136
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.72120369339109,
+      "learning_rate": 9.707530239973257e-06,
+      "loss": 0.5314,
+      "step": 2137
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6497313724067862,
+      "learning_rate": 9.707182499267851e-06,
+      "loss": 0.5609,
+      "step": 2138
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8843671026780996,
+      "learning_rate": 9.706834558192728e-06,
+      "loss": 0.6537,
+      "step": 2139
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6927955833608418,
+      "learning_rate": 9.706486416762696e-06,
+      "loss": 0.5397,
+      "step": 2140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7035186658685634,
+      "learning_rate": 9.706138074992581e-06,
+      "loss": 0.625,
+      "step": 2141
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6516548353517828,
+      "learning_rate": 9.705789532897205e-06,
+      "loss": 0.5345,
+      "step": 2142
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6039563025733024,
+      "learning_rate": 9.705440790491406e-06,
+      "loss": 0.5231,
+      "step": 2143
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7199901116167109,
+      "learning_rate": 9.705091847790029e-06,
+      "loss": 0.6729,
+      "step": 2144
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8138226573688557,
+      "learning_rate": 9.704742704807928e-06,
+      "loss": 0.6693,
+      "step": 2145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8791064321004185,
+      "learning_rate": 9.704393361559963e-06,
+      "loss": 0.5696,
+      "step": 2146
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7829314545257753,
+      "learning_rate": 9.704043818061007e-06,
+      "loss": 0.5571,
+      "step": 2147
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9699664155305645,
+      "learning_rate": 9.703694074325935e-06,
+      "loss": 0.6281,
+      "step": 2148
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8313151676250713,
+      "learning_rate": 9.70334413036964e-06,
+      "loss": 0.6658,
+      "step": 2149
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6751066000361966,
+      "learning_rate": 9.702993986207014e-06,
+      "loss": 0.5604,
+      "step": 2150
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7255453828137651,
+      "learning_rate": 9.702643641852963e-06,
+      "loss": 0.5996,
+      "step": 2151
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6983987210473735,
+      "learning_rate": 9.7022930973224e-06,
+      "loss": 0.5415,
+      "step": 2152
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.936193879001611,
+      "learning_rate": 9.701942352630246e-06,
+      "loss": 0.6409,
+      "step": 2153
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.912956042654097,
+      "learning_rate": 9.701591407791431e-06,
+      "loss": 0.6389,
+      "step": 2154
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7230062589302516,
+      "learning_rate": 9.701240262820894e-06,
+      "loss": 0.5882,
+      "step": 2155
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6272954988725057,
+      "learning_rate": 9.700888917733582e-06,
+      "loss": 0.5692,
+      "step": 2156
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.322684260098865,
+      "learning_rate": 9.70053737254445e-06,
+      "loss": 0.6119,
+      "step": 2157
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6771476018489737,
+      "learning_rate": 9.700185627268463e-06,
+      "loss": 0.5113,
+      "step": 2158
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6986586269348238,
+      "learning_rate": 9.699833681920595e-06,
+      "loss": 0.5369,
+      "step": 2159
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8707171763834329,
+      "learning_rate": 9.699481536515824e-06,
+      "loss": 0.6593,
+      "step": 2160
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0172857832492763,
+      "learning_rate": 9.699129191069141e-06,
+      "loss": 0.6357,
+      "step": 2161
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7348842636230057,
+      "learning_rate": 9.698776645595546e-06,
+      "loss": 0.6248,
+      "step": 2162
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.798755867497702,
+      "learning_rate": 9.698423900110046e-06,
+      "loss": 0.6579,
+      "step": 2163
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1928872004184816,
+      "learning_rate": 9.698070954627652e-06,
+      "loss": 0.664,
+      "step": 2164
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7014593433875881,
+      "learning_rate": 9.69771780916339e-06,
+      "loss": 0.5624,
+      "step": 2165
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.767277100934089,
+      "learning_rate": 9.697364463732293e-06,
+      "loss": 0.6384,
+      "step": 2166
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9248346711069855,
+      "learning_rate": 9.697010918349402e-06,
+      "loss": 0.6596,
+      "step": 2167
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6387516173536527,
+      "learning_rate": 9.696657173029767e-06,
+      "loss": 0.5607,
+      "step": 2168
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8047509435201944,
+      "learning_rate": 9.696303227788441e-06,
+      "loss": 0.6464,
+      "step": 2169
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6238519105509818,
+      "learning_rate": 9.695949082640497e-06,
+      "loss": 0.5011,
+      "step": 2170
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7983154023591998,
+      "learning_rate": 9.695594737601006e-06,
+      "loss": 0.5157,
+      "step": 2171
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.948630607940082,
+      "learning_rate": 9.695240192685051e-06,
+      "loss": 0.6251,
+      "step": 2172
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5868660188042967,
+      "learning_rate": 9.694885447907726e-06,
+      "loss": 0.5212,
+      "step": 2173
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7581979463662551,
+      "learning_rate": 9.69453050328413e-06,
+      "loss": 0.6261,
+      "step": 2174
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7541088329169049,
+      "learning_rate": 9.694175358829372e-06,
+      "loss": 0.6103,
+      "step": 2175
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5998178387441944,
+      "learning_rate": 9.693820014558568e-06,
+      "loss": 0.5542,
+      "step": 2176
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7863609600049947,
+      "learning_rate": 9.693464470486847e-06,
+      "loss": 0.6182,
+      "step": 2177
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6913994240424528,
+      "learning_rate": 9.69310872662934e-06,
+      "loss": 0.6005,
+      "step": 2178
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9020903523450176,
+      "learning_rate": 9.692752783001194e-06,
+      "loss": 0.6919,
+      "step": 2179
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8009512130787827,
+      "learning_rate": 9.692396639617556e-06,
+      "loss": 0.6028,
+      "step": 2180
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5603008308784059,
+      "learning_rate": 9.692040296493587e-06,
+      "loss": 0.5243,
+      "step": 2181
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.804830335848704,
+      "learning_rate": 9.691683753644459e-06,
+      "loss": 0.5833,
+      "step": 2182
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7136042371926369,
+      "learning_rate": 9.691327011085344e-06,
+      "loss": 0.5521,
+      "step": 2183
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7516852237198544,
+      "learning_rate": 9.690970068831431e-06,
+      "loss": 0.6193,
+      "step": 2184
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6088908958862602,
+      "learning_rate": 9.690612926897911e-06,
+      "loss": 0.5562,
+      "step": 2185
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7406939704811897,
+      "learning_rate": 9.690255585299988e-06,
+      "loss": 0.6627,
+      "step": 2186
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9702544059608751,
+      "learning_rate": 9.689898044052872e-06,
+      "loss": 0.6357,
+      "step": 2187
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6700948045728238,
+      "learning_rate": 9.689540303171785e-06,
+      "loss": 0.5431,
+      "step": 2188
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7436438109257585,
+      "learning_rate": 9.689182362671952e-06,
+      "loss": 0.6278,
+      "step": 2189
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7576916423630417,
+      "learning_rate": 9.688824222568608e-06,
+      "loss": 0.6341,
+      "step": 2190
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6391553504785336,
+      "learning_rate": 9.688465882877005e-06,
+      "loss": 0.4978,
+      "step": 2191
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0998597734639064,
+      "learning_rate": 9.688107343612387e-06,
+      "loss": 0.6933,
+      "step": 2192
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6994534113107235,
+      "learning_rate": 9.687748604790024e-06,
+      "loss": 0.5918,
+      "step": 2193
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7935273500053455,
+      "learning_rate": 9.68738966642518e-06,
+      "loss": 0.5797,
+      "step": 2194
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6313576338974369,
+      "learning_rate": 9.687030528533137e-06,
+      "loss": 0.5135,
+      "step": 2195
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.58855144835993,
+      "learning_rate": 9.686671191129184e-06,
+      "loss": 0.564,
+      "step": 2196
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6397217492402749,
+      "learning_rate": 9.686311654228613e-06,
+      "loss": 0.5565,
+      "step": 2197
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6257870798373377,
+      "learning_rate": 9.68595191784673e-06,
+      "loss": 0.5277,
+      "step": 2198
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7337667986495767,
+      "learning_rate": 9.68559198199885e-06,
+      "loss": 0.6309,
+      "step": 2199
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5983308697317115,
+      "learning_rate": 9.685231846700292e-06,
+      "loss": 0.5012,
+      "step": 2200
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6799392679002891,
+      "learning_rate": 9.684871511966383e-06,
+      "loss": 0.5181,
+      "step": 2201
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6914395568735426,
+      "learning_rate": 9.684510977812467e-06,
+      "loss": 0.5455,
+      "step": 2202
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6238811969387991,
+      "learning_rate": 9.68415024425389e-06,
+      "loss": 0.5432,
+      "step": 2203
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6340758968912651,
+      "learning_rate": 9.683789311306003e-06,
+      "loss": 0.5409,
+      "step": 2204
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6886613532073099,
+      "learning_rate": 9.683428178984172e-06,
+      "loss": 0.5815,
+      "step": 2205
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6442929205541311,
+      "learning_rate": 9.68306684730377e-06,
+      "loss": 0.5504,
+      "step": 2206
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8588064591456216,
+      "learning_rate": 9.682705316280178e-06,
+      "loss": 0.6147,
+      "step": 2207
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8804960211736516,
+      "learning_rate": 9.682343585928785e-06,
+      "loss": 0.6474,
+      "step": 2208
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9515813480598867,
+      "learning_rate": 9.681981656264984e-06,
+      "loss": 0.6407,
+      "step": 2209
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5990587431565297,
+      "learning_rate": 9.68161952730419e-06,
+      "loss": 0.5491,
+      "step": 2210
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6515404033001146,
+      "learning_rate": 9.681257199061812e-06,
+      "loss": 0.5397,
+      "step": 2211
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5903850281832909,
+      "learning_rate": 9.680894671553273e-06,
+      "loss": 0.5919,
+      "step": 2212
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7799265947403901,
+      "learning_rate": 9.680531944794007e-06,
+      "loss": 0.6372,
+      "step": 2213
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9926507262328429,
+      "learning_rate": 9.680169018799452e-06,
+      "loss": 0.5947,
+      "step": 2214
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6339056599829482,
+      "learning_rate": 9.679805893585059e-06,
+      "loss": 0.4961,
+      "step": 2215
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6551315532687673,
+      "learning_rate": 9.679442569166282e-06,
+      "loss": 0.5736,
+      "step": 2216
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6300361657396402,
+      "learning_rate": 9.67907904555859e-06,
+      "loss": 0.5434,
+      "step": 2217
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9418706930656059,
+      "learning_rate": 9.678715322777453e-06,
+      "loss": 0.6208,
+      "step": 2218
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6543659916555796,
+      "learning_rate": 9.678351400838357e-06,
+      "loss": 0.5593,
+      "step": 2219
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8837889611829904,
+      "learning_rate": 9.67798727975679e-06,
+      "loss": 0.7165,
+      "step": 2220
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8815426881808548,
+      "learning_rate": 9.677622959548256e-06,
+      "loss": 0.6614,
+      "step": 2221
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6174900120372893,
+      "learning_rate": 9.677258440228259e-06,
+      "loss": 0.5333,
+      "step": 2222
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8184534973285272,
+      "learning_rate": 9.676893721812317e-06,
+      "loss": 0.6351,
+      "step": 2223
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6330019368009169,
+      "learning_rate": 9.676528804315954e-06,
+      "loss": 0.5142,
+      "step": 2224
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8743920706670454,
+      "learning_rate": 9.676163687754705e-06,
+      "loss": 0.6303,
+      "step": 2225
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6326861851605023,
+      "learning_rate": 9.675798372144108e-06,
+      "loss": 0.5446,
+      "step": 2226
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0371082516296481,
+      "learning_rate": 9.675432857499718e-06,
+      "loss": 0.5686,
+      "step": 2227
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.836242150354895,
+      "learning_rate": 9.675067143837092e-06,
+      "loss": 0.642,
+      "step": 2228
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8301236096375209,
+      "learning_rate": 9.674701231171795e-06,
+      "loss": 0.5979,
+      "step": 2229
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1578010235586267,
+      "learning_rate": 9.674335119519407e-06,
+      "loss": 0.6818,
+      "step": 2230
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9290913461612924,
+      "learning_rate": 9.673968808895509e-06,
+      "loss": 0.6093,
+      "step": 2231
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7342519929147039,
+      "learning_rate": 9.673602299315694e-06,
+      "loss": 0.596,
+      "step": 2232
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6525783827942293,
+      "learning_rate": 9.673235590795565e-06,
+      "loss": 0.5375,
+      "step": 2233
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5989832941113682,
+      "learning_rate": 9.672868683350731e-06,
+      "loss": 0.5109,
+      "step": 2234
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6569948231952824,
+      "learning_rate": 9.67250157699681e-06,
+      "loss": 0.5465,
+      "step": 2235
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6724503094389964,
+      "learning_rate": 9.672134271749425e-06,
+      "loss": 0.5342,
+      "step": 2236
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0433338392795097,
+      "learning_rate": 9.671766767624215e-06,
+      "loss": 0.659,
+      "step": 2237
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6852954531542389,
+      "learning_rate": 9.671399064636824e-06,
+      "loss": 0.5177,
+      "step": 2238
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7620657504708801,
+      "learning_rate": 9.671031162802901e-06,
+      "loss": 0.6211,
+      "step": 2239
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7714690157358612,
+      "learning_rate": 9.670663062138111e-06,
+      "loss": 0.5459,
+      "step": 2240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6553942130009848,
+      "learning_rate": 9.670294762658116e-06,
+      "loss": 0.546,
+      "step": 2241
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.62564941691109,
+      "learning_rate": 9.669926264378598e-06,
+      "loss": 0.556,
+      "step": 2242
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5762097403614382,
+      "learning_rate": 9.669557567315242e-06,
+      "loss": 0.5195,
+      "step": 2243
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8532764287259377,
+      "learning_rate": 9.669188671483742e-06,
+      "loss": 0.6213,
+      "step": 2244
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7489377850365356,
+      "learning_rate": 9.668819576899802e-06,
+      "loss": 0.5236,
+      "step": 2245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7617188167186214,
+      "learning_rate": 9.668450283579132e-06,
+      "loss": 0.6345,
+      "step": 2246
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.851474713301415,
+      "learning_rate": 9.668080791537451e-06,
+      "loss": 0.6059,
+      "step": 2247
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7362806713736375,
+      "learning_rate": 9.667711100790487e-06,
+      "loss": 0.5474,
+      "step": 2248
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6993209279475131,
+      "learning_rate": 9.667341211353979e-06,
+      "loss": 0.5467,
+      "step": 2249
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6047618263812696,
+      "learning_rate": 9.66697112324367e-06,
+      "loss": 0.5799,
+      "step": 2250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6760460559516863,
+      "learning_rate": 9.666600836475313e-06,
+      "loss": 0.5667,
+      "step": 2251
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8904894784357311,
+      "learning_rate": 9.66623035106467e-06,
+      "loss": 0.6271,
+      "step": 2252
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5959552961352129,
+      "learning_rate": 9.665859667027514e-06,
+      "loss": 0.5595,
+      "step": 2253
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5915119580046773,
+      "learning_rate": 9.665488784379619e-06,
+      "loss": 0.5449,
+      "step": 2254
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7050973527303205,
+      "learning_rate": 9.665117703136778e-06,
+      "loss": 0.5475,
+      "step": 2255
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5741843191594764,
+      "learning_rate": 9.664746423314783e-06,
+      "loss": 0.4807,
+      "step": 2256
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.605516106318389,
+      "learning_rate": 9.66437494492944e-06,
+      "loss": 0.4949,
+      "step": 2257
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7298421166590036,
+      "learning_rate": 9.66400326799656e-06,
+      "loss": 0.6303,
+      "step": 2258
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8276973612240265,
+      "learning_rate": 9.663631392531964e-06,
+      "loss": 0.5599,
+      "step": 2259
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7930071135800971,
+      "learning_rate": 9.663259318551484e-06,
+      "loss": 0.5948,
+      "step": 2260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8995766276937424,
+      "learning_rate": 9.662887046070955e-06,
+      "loss": 0.5952,
+      "step": 2261
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6427601142691367,
+      "learning_rate": 9.662514575106226e-06,
+      "loss": 0.5744,
+      "step": 2262
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.828253738518374,
+      "learning_rate": 9.66214190567315e-06,
+      "loss": 0.6222,
+      "step": 2263
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7720476783409141,
+      "learning_rate": 9.661769037787593e-06,
+      "loss": 0.5892,
+      "step": 2264
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5936003716856899,
+      "learning_rate": 9.661395971465425e-06,
+      "loss": 0.4931,
+      "step": 2265
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6928100126960826,
+      "learning_rate": 9.661022706722522e-06,
+      "loss": 0.5143,
+      "step": 2266
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6655182767575827,
+      "learning_rate": 9.66064924357478e-06,
+      "loss": 0.5259,
+      "step": 2267
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7407844825951722,
+      "learning_rate": 9.660275582038095e-06,
+      "loss": 0.5317,
+      "step": 2268
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7158721229727145,
+      "learning_rate": 9.659901722128366e-06,
+      "loss": 0.5526,
+      "step": 2269
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7424287306296785,
+      "learning_rate": 9.659527663861513e-06,
+      "loss": 0.6317,
+      "step": 2270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.107374494985659,
+      "learning_rate": 9.65915340725346e-06,
+      "loss": 0.6258,
+      "step": 2271
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6830529175884509,
+      "learning_rate": 9.658778952320133e-06,
+      "loss": 0.6194,
+      "step": 2272
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6713118487544459,
+      "learning_rate": 9.658404299077472e-06,
+      "loss": 0.5113,
+      "step": 2273
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8508721732925933,
+      "learning_rate": 9.658029447541429e-06,
+      "loss": 0.5819,
+      "step": 2274
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6838730161706225,
+      "learning_rate": 9.657654397727956e-06,
+      "loss": 0.5601,
+      "step": 2275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7911712009038191,
+      "learning_rate": 9.657279149653018e-06,
+      "loss": 0.5822,
+      "step": 2276
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.776451702863992,
+      "learning_rate": 9.65690370333259e-06,
+      "loss": 0.6277,
+      "step": 2277
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8645971786151312,
+      "learning_rate": 9.656528058782653e-06,
+      "loss": 0.6251,
+      "step": 2278
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8527749737790213,
+      "learning_rate": 9.656152216019197e-06,
+      "loss": 0.6788,
+      "step": 2279
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8828743894008848,
+      "learning_rate": 9.655776175058218e-06,
+      "loss": 0.6028,
+      "step": 2280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.823956024995578,
+      "learning_rate": 9.655399935915728e-06,
+      "loss": 0.6078,
+      "step": 2281
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.925549441991507,
+      "learning_rate": 9.655023498607736e-06,
+      "loss": 0.6605,
+      "step": 2282
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6801753009251831,
+      "learning_rate": 9.654646863150271e-06,
+      "loss": 0.568,
+      "step": 2283
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7717913263725459,
+      "learning_rate": 9.654270029559362e-06,
+      "loss": 0.6152,
+      "step": 2284
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8557536389595195,
+      "learning_rate": 9.653892997851052e-06,
+      "loss": 0.6703,
+      "step": 2285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8793890657596471,
+      "learning_rate": 9.653515768041388e-06,
+      "loss": 0.6016,
+      "step": 2286
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6443234170989017,
+      "learning_rate": 9.653138340146429e-06,
+      "loss": 0.5274,
+      "step": 2287
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8172794242502196,
+      "learning_rate": 9.65276071418224e-06,
+      "loss": 0.5886,
+      "step": 2288
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.686034579986197,
+      "learning_rate": 9.652382890164895e-06,
+      "loss": 0.5846,
+      "step": 2289
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8096073641913435,
+      "learning_rate": 9.652004868110477e-06,
+      "loss": 0.5808,
+      "step": 2290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.624890273928074,
+      "learning_rate": 9.65162664803508e-06,
+      "loss": 0.5363,
+      "step": 2291
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6749908950627436,
+      "learning_rate": 9.651248229954798e-06,
+      "loss": 0.6828,
+      "step": 2292
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8880591936931553,
+      "learning_rate": 9.650869613885742e-06,
+      "loss": 0.6491,
+      "step": 2293
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8159750356945049,
+      "learning_rate": 9.65049079984403e-06,
+      "loss": 0.6944,
+      "step": 2294
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7418131176842386,
+      "learning_rate": 9.650111787845784e-06,
+      "loss": 0.6,
+      "step": 2295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0461700823998974,
+      "learning_rate": 9.64973257790714e-06,
+      "loss": 0.5587,
+      "step": 2296
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6101097244691224,
+      "learning_rate": 9.64935317004424e-06,
+      "loss": 0.5083,
+      "step": 2297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.275132176881301,
+      "learning_rate": 9.648973564273232e-06,
+      "loss": 0.5649,
+      "step": 2298
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7032765826619104,
+      "learning_rate": 9.648593760610274e-06,
+      "loss": 0.6405,
+      "step": 2299
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6759865672860612,
+      "learning_rate": 9.648213759071536e-06,
+      "loss": 0.4923,
+      "step": 2300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8516325001087683,
+      "learning_rate": 9.64783355967319e-06,
+      "loss": 0.5733,
+      "step": 2301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0391152789607287,
+      "learning_rate": 9.647453162431422e-06,
+      "loss": 0.6847,
+      "step": 2302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7283393465124091,
+      "learning_rate": 9.647072567362426e-06,
+      "loss": 0.6249,
+      "step": 2303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5795741497487102,
+      "learning_rate": 9.6466917744824e-06,
+      "loss": 0.548,
+      "step": 2304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6297529108765231,
+      "learning_rate": 9.646310783807552e-06,
+      "loss": 0.5269,
+      "step": 2305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7688890632750403,
+      "learning_rate": 9.645929595354101e-06,
+      "loss": 0.583,
+      "step": 2306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5647579057091476,
+      "learning_rate": 9.645548209138277e-06,
+      "loss": 0.4726,
+      "step": 2307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6696597283297824,
+      "learning_rate": 9.645166625176308e-06,
+      "loss": 0.5587,
+      "step": 2308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5912881255480464,
+      "learning_rate": 9.64478484348444e-06,
+      "loss": 0.5827,
+      "step": 2309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9013934620363223,
+      "learning_rate": 9.644402864078925e-06,
+      "loss": 0.6017,
+      "step": 2310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8091548399589423,
+      "learning_rate": 9.644020686976022e-06,
+      "loss": 0.5655,
+      "step": 2311
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6815593571986776,
+      "learning_rate": 9.643638312191996e-06,
+      "loss": 0.5179,
+      "step": 2312
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6353061674175475,
+      "learning_rate": 9.643255739743128e-06,
+      "loss": 0.5628,
+      "step": 2313
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5980690191909007,
+      "learning_rate": 9.6428729696457e-06,
+      "loss": 0.5004,
+      "step": 2314
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5907665810638241,
+      "learning_rate": 9.642490001916004e-06,
+      "loss": 0.584,
+      "step": 2315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6795787447301739,
+      "learning_rate": 9.642106836570349e-06,
+      "loss": 0.5571,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6581148947684329,
+      "learning_rate": 9.641723473625036e-06,
+      "loss": 0.5441,
+      "step": 2317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6025978028439085,
+      "learning_rate": 9.64133991309639e-06,
+      "loss": 0.5222,
+      "step": 2318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6102422844649658,
+      "learning_rate": 9.640956155000734e-06,
+      "loss": 0.4969,
+      "step": 2319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5951995875521914,
+      "learning_rate": 9.640572199354404e-06,
+      "loss": 0.4884,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8117996206937678,
+      "learning_rate": 9.640188046173746e-06,
+      "loss": 0.6194,
+      "step": 2321
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.569777508360132,
+      "learning_rate": 9.639803695475111e-06,
+      "loss": 0.501,
+      "step": 2322
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7353086269939838,
+      "learning_rate": 9.639419147274858e-06,
+      "loss": 0.5792,
+      "step": 2323
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8305030844985611,
+      "learning_rate": 9.639034401589359e-06,
+      "loss": 0.6425,
+      "step": 2324
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6875012573272089,
+      "learning_rate": 9.638649458434989e-06,
+      "loss": 0.4913,
+      "step": 2325
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6104260420512554,
+      "learning_rate": 9.638264317828135e-06,
+      "loss": 0.5186,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8513012735792077,
+      "learning_rate": 9.63787897978519e-06,
+      "loss": 0.6214,
+      "step": 2327
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8140359431336607,
+      "learning_rate": 9.637493444322557e-06,
+      "loss": 0.6138,
+      "step": 2328
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9359810761278444,
+      "learning_rate": 9.63710771145665e-06,
+      "loss": 0.5982,
+      "step": 2329
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7033344549949977,
+      "learning_rate": 9.636721781203882e-06,
+      "loss": 0.5665,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.712401301102582,
+      "learning_rate": 9.636335653580687e-06,
+      "loss": 0.5778,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6295356997241944,
+      "learning_rate": 9.635949328603499e-06,
+      "loss": 0.5391,
+      "step": 2332
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7082610788110011,
+      "learning_rate": 9.635562806288763e-06,
+      "loss": 0.6003,
+      "step": 2333
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5721241947970244,
+      "learning_rate": 9.635176086652929e-06,
+      "loss": 0.5039,
+      "step": 2334
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3745115336996983,
+      "learning_rate": 9.634789169712461e-06,
+      "loss": 0.6207,
+      "step": 2335
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9141369306134102,
+      "learning_rate": 9.634402055483832e-06,
+      "loss": 0.608,
+      "step": 2336
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6391256201758749,
+      "learning_rate": 9.634014743983513e-06,
+      "loss": 0.548,
+      "step": 2337
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6602876220896754,
+      "learning_rate": 9.633627235227998e-06,
+      "loss": 0.5919,
+      "step": 2338
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7155167772219896,
+      "learning_rate": 9.633239529233776e-06,
+      "loss": 0.4985,
+      "step": 2339
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7546479779069812,
+      "learning_rate": 9.632851626017355e-06,
+      "loss": 0.6248,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9036446272415479,
+      "learning_rate": 9.632463525595243e-06,
+      "loss": 0.6682,
+      "step": 2341
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.692489419858691,
+      "learning_rate": 9.632075227983963e-06,
+      "loss": 0.5552,
+      "step": 2342
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6542026424178569,
+      "learning_rate": 9.63168673320004e-06,
+      "loss": 0.5592,
+      "step": 2343
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9882606063086841,
+      "learning_rate": 9.631298041260018e-06,
+      "loss": 0.6924,
+      "step": 2344
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.102230759389261,
+      "learning_rate": 9.630909152180434e-06,
+      "loss": 0.6162,
+      "step": 2345
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7644577021931153,
+      "learning_rate": 9.63052006597785e-06,
+      "loss": 0.5817,
+      "step": 2346
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.716617409334602,
+      "learning_rate": 9.630130782668818e-06,
+      "loss": 0.5892,
+      "step": 2347
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9729341324652461,
+      "learning_rate": 9.629741302269918e-06,
+      "loss": 0.6403,
+      "step": 2348
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6853664457080397,
+      "learning_rate": 9.629351624797725e-06,
+      "loss": 0.6067,
+      "step": 2349
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8342759285622677,
+      "learning_rate": 9.628961750268825e-06,
+      "loss": 0.632,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7689119065965148,
+      "learning_rate": 9.628571678699818e-06,
+      "loss": 0.573,
+      "step": 2351
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.858385414084931,
+      "learning_rate": 9.628181410107305e-06,
+      "loss": 0.7072,
+      "step": 2352
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7249149251190334,
+      "learning_rate": 9.627790944507898e-06,
+      "loss": 0.5645,
+      "step": 2353
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6217419898343416,
+      "learning_rate": 9.627400281918218e-06,
+      "loss": 0.5539,
+      "step": 2354
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7610052903814447,
+      "learning_rate": 9.627009422354896e-06,
+      "loss": 0.6443,
+      "step": 2355
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9250303565161557,
+      "learning_rate": 9.626618365834568e-06,
+      "loss": 0.5487,
+      "step": 2356
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8242894584329374,
+      "learning_rate": 9.62622711237388e-06,
+      "loss": 0.6333,
+      "step": 2357
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8640955393357904,
+      "learning_rate": 9.62583566198949e-06,
+      "loss": 0.6112,
+      "step": 2358
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5736128865275786,
+      "learning_rate": 9.625444014698056e-06,
+      "loss": 0.5134,
+      "step": 2359
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8840592830174363,
+      "learning_rate": 9.62505217051625e-06,
+      "loss": 0.6405,
+      "step": 2360
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.782897049526451,
+      "learning_rate": 9.624660129460756e-06,
+      "loss": 0.6289,
+      "step": 2361
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7876266615541588,
+      "learning_rate": 9.624267891548257e-06,
+      "loss": 0.6221,
+      "step": 2362
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9185263520658862,
+      "learning_rate": 9.62387545679545e-06,
+      "loss": 0.691,
+      "step": 2363
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6542667272945265,
+      "learning_rate": 9.623482825219041e-06,
+      "loss": 0.5247,
+      "step": 2364
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6180807685369759,
+      "learning_rate": 9.623089996835744e-06,
+      "loss": 0.5468,
+      "step": 2365
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.713283927268299,
+      "learning_rate": 9.622696971662278e-06,
+      "loss": 0.6307,
+      "step": 2366
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.87003445307037,
+      "learning_rate": 9.622303749715375e-06,
+      "loss": 0.6911,
+      "step": 2367
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.94623684327798,
+      "learning_rate": 9.621910331011769e-06,
+      "loss": 0.6401,
+      "step": 2368
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7600855773772919,
+      "learning_rate": 9.621516715568212e-06,
+      "loss": 0.5957,
+      "step": 2369
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.874117530399109,
+      "learning_rate": 9.621122903401457e-06,
+      "loss": 0.6094,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8481734501936808,
+      "learning_rate": 9.620728894528266e-06,
+      "loss": 0.6145,
+      "step": 2371
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7617368364319198,
+      "learning_rate": 9.620334688965411e-06,
+      "loss": 0.6298,
+      "step": 2372
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7728250496801723,
+      "learning_rate": 9.619940286729674e-06,
+      "loss": 0.5981,
+      "step": 2373
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6006432571031343,
+      "learning_rate": 9.619545687837843e-06,
+      "loss": 0.5815,
+      "step": 2374
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7252849051821438,
+      "learning_rate": 9.619150892306713e-06,
+      "loss": 0.6159,
+      "step": 2375
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6607678635799454,
+      "learning_rate": 9.618755900153091e-06,
+      "loss": 0.5698,
+      "step": 2376
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.594289771715513,
+      "learning_rate": 9.618360711393789e-06,
+      "loss": 0.5386,
+      "step": 2377
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6907339889550254,
+      "learning_rate": 9.61796532604563e-06,
+      "loss": 0.57,
+      "step": 2378
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6174250140212003,
+      "learning_rate": 9.617569744125443e-06,
+      "loss": 0.5633,
+      "step": 2379
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8122324273378344,
+      "learning_rate": 9.617173965650068e-06,
+      "loss": 0.6107,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.092650696850718,
+      "learning_rate": 9.616777990636353e-06,
+      "loss": 0.6218,
+      "step": 2381
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6128014853222221,
+      "learning_rate": 9.616381819101151e-06,
+      "loss": 0.5706,
+      "step": 2382
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5657710730165172,
+      "learning_rate": 9.615985451061327e-06,
+      "loss": 0.4872,
+      "step": 2383
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9024634261132601,
+      "learning_rate": 9.615588886533753e-06,
+      "loss": 0.6439,
+      "step": 2384
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6598051106064764,
+      "learning_rate": 9.615192125535308e-06,
+      "loss": 0.5426,
+      "step": 2385
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.941218753522489,
+      "learning_rate": 9.614795168082885e-06,
+      "loss": 0.6302,
+      "step": 2386
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6969140643920829,
+      "learning_rate": 9.614398014193376e-06,
+      "loss": 0.5691,
+      "step": 2387
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6307159750972525,
+      "learning_rate": 9.614000663883692e-06,
+      "loss": 0.5168,
+      "step": 2388
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7211489461599196,
+      "learning_rate": 9.61360311717074e-06,
+      "loss": 0.6418,
+      "step": 2389
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6542520301640792,
+      "learning_rate": 9.613205374071449e-06,
+      "loss": 0.5681,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7676464715721771,
+      "learning_rate": 9.612807434602747e-06,
+      "loss": 0.567,
+      "step": 2391
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6810222458656829,
+      "learning_rate": 9.612409298781571e-06,
+      "loss": 0.5391,
+      "step": 2392
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6421619680166654,
+      "learning_rate": 9.612010966624871e-06,
+      "loss": 0.5632,
+      "step": 2393
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6334382080486921,
+      "learning_rate": 9.611612438149604e-06,
+      "loss": 0.5109,
+      "step": 2394
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6000212034183585,
+      "learning_rate": 9.61121371337273e-06,
+      "loss": 0.5949,
+      "step": 2395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6461431312045264,
+      "learning_rate": 9.610814792311223e-06,
+      "loss": 0.5368,
+      "step": 2396
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7603082067529316,
+      "learning_rate": 9.610415674982066e-06,
+      "loss": 0.6449,
+      "step": 2397
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8991449463517307,
+      "learning_rate": 9.610016361402246e-06,
+      "loss": 0.6748,
+      "step": 2398
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6349684387182754,
+      "learning_rate": 9.60961685158876e-06,
+      "loss": 0.5669,
+      "step": 2399
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6248834326375421,
+      "learning_rate": 9.609217145558617e-06,
+      "loss": 0.5188,
+      "step": 2400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2104698116864234,
+      "learning_rate": 9.608817243328827e-06,
+      "loss": 0.7036,
+      "step": 2401
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8889571647654392,
+      "learning_rate": 9.608417144916417e-06,
+      "loss": 0.6893,
+      "step": 2402
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6054003487784059,
+      "learning_rate": 9.608016850338413e-06,
+      "loss": 0.5112,
+      "step": 2403
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.704592851394262,
+      "learning_rate": 9.60761635961186e-06,
+      "loss": 0.5493,
+      "step": 2404
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.551236892913067,
+      "learning_rate": 9.607215672753799e-06,
+      "loss": 0.4779,
+      "step": 2405
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.552987497017455,
+      "learning_rate": 9.60681478978129e-06,
+      "loss": 0.5193,
+      "step": 2406
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5834679274972107,
+      "learning_rate": 9.606413710711398e-06,
+      "loss": 0.5557,
+      "step": 2407
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8429905180044114,
+      "learning_rate": 9.606012435561194e-06,
+      "loss": 0.6261,
+      "step": 2408
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5815754742957093,
+      "learning_rate": 9.605610964347758e-06,
+      "loss": 0.5468,
+      "step": 2409
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5785602730013256,
+      "learning_rate": 9.605209297088182e-06,
+      "loss": 0.546,
+      "step": 2410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7059173564552337,
+      "learning_rate": 9.604807433799563e-06,
+      "loss": 0.583,
+      "step": 2411
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5967318636854467,
+      "learning_rate": 9.604405374499003e-06,
+      "loss": 0.552,
+      "step": 2412
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7672505433207766,
+      "learning_rate": 9.604003119203624e-06,
+      "loss": 0.5824,
+      "step": 2413
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6569396385308869,
+      "learning_rate": 9.603600667930542e-06,
+      "loss": 0.5255,
+      "step": 2414
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7784896520611584,
+      "learning_rate": 9.603198020696892e-06,
+      "loss": 0.5965,
+      "step": 2415
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5673147912890423,
+      "learning_rate": 9.60279517751981e-06,
+      "loss": 0.497,
+      "step": 2416
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6463402705649699,
+      "learning_rate": 9.602392138416447e-06,
+      "loss": 0.5272,
+      "step": 2417
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6907065927904141,
+      "learning_rate": 9.601988903403958e-06,
+      "loss": 0.5938,
+      "step": 2418
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8793334566508969,
+      "learning_rate": 9.601585472499508e-06,
+      "loss": 0.6642,
+      "step": 2419
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7844761009137756,
+      "learning_rate": 9.601181845720268e-06,
+      "loss": 0.5851,
+      "step": 2420
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.67963826962421,
+      "learning_rate": 9.60077802308342e-06,
+      "loss": 0.5639,
+      "step": 2421
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8693147114331978,
+      "learning_rate": 9.600374004606153e-06,
+      "loss": 0.5973,
+      "step": 2422
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6078543400729967,
+      "learning_rate": 9.599969790305667e-06,
+      "loss": 0.5417,
+      "step": 2423
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8805026740278471,
+      "learning_rate": 9.599565380199164e-06,
+      "loss": 0.6367,
+      "step": 2424
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5795579343765204,
+      "learning_rate": 9.599160774303863e-06,
+      "loss": 0.505,
+      "step": 2425
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8687021496402021,
+      "learning_rate": 9.598755972636983e-06,
+      "loss": 0.6368,
+      "step": 2426
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7197724942923636,
+      "learning_rate": 9.598350975215757e-06,
+      "loss": 0.5701,
+      "step": 2427
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.723969890496662,
+      "learning_rate": 9.597945782057427e-06,
+      "loss": 0.6485,
+      "step": 2428
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7725077900043601,
+      "learning_rate": 9.597540393179235e-06,
+      "loss": 0.6412,
+      "step": 2429
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8068132619636262,
+      "learning_rate": 9.59713480859844e-06,
+      "loss": 0.7052,
+      "step": 2430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.624133680104771,
+      "learning_rate": 9.596729028332309e-06,
+      "loss": 0.5699,
+      "step": 2431
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7688880035065165,
+      "learning_rate": 9.596323052398112e-06,
+      "loss": 0.6272,
+      "step": 2432
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8384046807259603,
+      "learning_rate": 9.595916880813127e-06,
+      "loss": 0.6034,
+      "step": 2433
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.769858270127807,
+      "learning_rate": 9.59551051359465e-06,
+      "loss": 0.5621,
+      "step": 2434
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6400763534881783,
+      "learning_rate": 9.595103950759974e-06,
+      "loss": 0.5397,
+      "step": 2435
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8557673589892321,
+      "learning_rate": 9.594697192326408e-06,
+      "loss": 0.5849,
+      "step": 2436
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6461724804678096,
+      "learning_rate": 9.594290238311264e-06,
+      "loss": 0.5297,
+      "step": 2437
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9378438918064688,
+      "learning_rate": 9.593883088731866e-06,
+      "loss": 0.6056,
+      "step": 2438
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8683897724391,
+      "learning_rate": 9.593475743605546e-06,
+      "loss": 0.6301,
+      "step": 2439
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8675075549901747,
+      "learning_rate": 9.593068202949642e-06,
+      "loss": 0.6534,
+      "step": 2440
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6497677519019572,
+      "learning_rate": 9.5926604667815e-06,
+      "loss": 0.5513,
+      "step": 2441
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6224714071357573,
+      "learning_rate": 9.59225253511848e-06,
+      "loss": 0.5239,
+      "step": 2442
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8110296456715863,
+      "learning_rate": 9.591844407977944e-06,
+      "loss": 0.6376,
+      "step": 2443
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5785413804034262,
+      "learning_rate": 9.591436085377263e-06,
+      "loss": 0.5523,
+      "step": 2444
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9300910909493508,
+      "learning_rate": 9.591027567333822e-06,
+      "loss": 0.6671,
+      "step": 2445
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7967115344511013,
+      "learning_rate": 9.590618853865008e-06,
+      "loss": 0.6202,
+      "step": 2446
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7293896493512114,
+      "learning_rate": 9.590209944988218e-06,
+      "loss": 0.5719,
+      "step": 2447
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6017138154585803,
+      "learning_rate": 9.58980084072086e-06,
+      "loss": 0.5371,
+      "step": 2448
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6763495265701278,
+      "learning_rate": 9.589391541080346e-06,
+      "loss": 0.5398,
+      "step": 2449
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.633634695181765,
+      "learning_rate": 9.588982046084101e-06,
+      "loss": 0.5232,
+      "step": 2450
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5540877292722046,
+      "learning_rate": 9.588572355749555e-06,
+      "loss": 0.5284,
+      "step": 2451
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9761110237084909,
+      "learning_rate": 9.588162470094145e-06,
+      "loss": 0.665,
+      "step": 2452
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.223077998431946,
+      "learning_rate": 9.58775238913532e-06,
+      "loss": 0.6697,
+      "step": 2453
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5979268632542212,
+      "learning_rate": 9.587342112890539e-06,
+      "loss": 0.5319,
+      "step": 2454
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7510362212559786,
+      "learning_rate": 9.586931641377262e-06,
+      "loss": 0.6268,
+      "step": 2455
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7176335642595973,
+      "learning_rate": 9.58652097461296e-06,
+      "loss": 0.5559,
+      "step": 2456
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6989847542259999,
+      "learning_rate": 9.58611011261512e-06,
+      "loss": 0.5358,
+      "step": 2457
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8864963162315552,
+      "learning_rate": 9.585699055401226e-06,
+      "loss": 0.6491,
+      "step": 2458
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6172125182878831,
+      "learning_rate": 9.585287802988778e-06,
+      "loss": 0.5386,
+      "step": 2459
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7234782866031692,
+      "learning_rate": 9.58487635539528e-06,
+      "loss": 0.5292,
+      "step": 2460
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6549233588633315,
+      "learning_rate": 9.584464712638245e-06,
+      "loss": 0.5821,
+      "step": 2461
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7295733671228533,
+      "learning_rate": 9.584052874735201e-06,
+      "loss": 0.6253,
+      "step": 2462
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5980296759480169,
+      "learning_rate": 9.583640841703672e-06,
+      "loss": 0.4917,
+      "step": 2463
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6224250812587412,
+      "learning_rate": 9.5832286135612e-06,
+      "loss": 0.5461,
+      "step": 2464
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6684864407034281,
+      "learning_rate": 9.582816190325333e-06,
+      "loss": 0.5555,
+      "step": 2465
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6233339430704379,
+      "learning_rate": 9.582403572013623e-06,
+      "loss": 0.5446,
+      "step": 2466
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7842371818556565,
+      "learning_rate": 9.58199075864364e-06,
+      "loss": 0.6205,
+      "step": 2467
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7908366264740696,
+      "learning_rate": 9.581577750232948e-06,
+      "loss": 0.5679,
+      "step": 2468
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7666704393633019,
+      "learning_rate": 9.581164546799135e-06,
+      "loss": 0.6375,
+      "step": 2469
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5828591565212466,
+      "learning_rate": 9.580751148359785e-06,
+      "loss": 0.5457,
+      "step": 2470
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6463208179192446,
+      "learning_rate": 9.580337554932497e-06,
+      "loss": 0.6051,
+      "step": 2471
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6733956994877681,
+      "learning_rate": 9.579923766534875e-06,
+      "loss": 0.5557,
+      "step": 2472
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7015158626325191,
+      "learning_rate": 9.579509783184535e-06,
+      "loss": 0.6169,
+      "step": 2473
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5563294339181992,
+      "learning_rate": 9.579095604899097e-06,
+      "loss": 0.5075,
+      "step": 2474
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9959362943126006,
+      "learning_rate": 9.578681231696191e-06,
+      "loss": 0.6288,
+      "step": 2475
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7535498086676388,
+      "learning_rate": 9.578266663593458e-06,
+      "loss": 0.7042,
+      "step": 2476
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.825513303560254,
+      "learning_rate": 9.577851900608541e-06,
+      "loss": 0.6261,
+      "step": 2477
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7635947749024156,
+      "learning_rate": 9.5774369427591e-06,
+      "loss": 0.5794,
+      "step": 2478
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5658271771333826,
+      "learning_rate": 9.577021790062794e-06,
+      "loss": 0.4903,
+      "step": 2479
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7353477955814989,
+      "learning_rate": 9.576606442537297e-06,
+      "loss": 0.5989,
+      "step": 2480
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7766002116136786,
+      "learning_rate": 9.576190900200288e-06,
+      "loss": 0.6137,
+      "step": 2481
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.889908014101899,
+      "learning_rate": 9.575775163069456e-06,
+      "loss": 0.6999,
+      "step": 2482
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9757343338918586,
+      "learning_rate": 9.575359231162497e-06,
+      "loss": 0.5649,
+      "step": 2483
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7849444011005536,
+      "learning_rate": 9.574943104497118e-06,
+      "loss": 0.6195,
+      "step": 2484
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6463259656534145,
+      "learning_rate": 9.574526783091029e-06,
+      "loss": 0.5361,
+      "step": 2485
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6617501805839066,
+      "learning_rate": 9.574110266961953e-06,
+      "loss": 0.5729,
+      "step": 2486
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7279993137004347,
+      "learning_rate": 9.573693556127618e-06,
+      "loss": 0.5579,
+      "step": 2487
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7996822044827662,
+      "learning_rate": 9.573276650605768e-06,
+      "loss": 0.6384,
+      "step": 2488
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5908133831685772,
+      "learning_rate": 9.572859550414143e-06,
+      "loss": 0.5495,
+      "step": 2489
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7043086949520345,
+      "learning_rate": 9.572442255570498e-06,
+      "loss": 0.5855,
+      "step": 2490
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8993169692043346,
+      "learning_rate": 9.5720247660926e-06,
+      "loss": 0.536,
+      "step": 2491
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7928167705587933,
+      "learning_rate": 9.571607081998216e-06,
+      "loss": 0.5852,
+      "step": 2492
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7593312595814335,
+      "learning_rate": 9.571189203305128e-06,
+      "loss": 0.6027,
+      "step": 2493
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6564232624986837,
+      "learning_rate": 9.570771130031123e-06,
+      "loss": 0.6041,
+      "step": 2494
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6221530793613381,
+      "learning_rate": 9.570352862193998e-06,
+      "loss": 0.5808,
+      "step": 2495
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7136315212099565,
+      "learning_rate": 9.569934399811556e-06,
+      "loss": 0.5321,
+      "step": 2496
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7963988455937888,
+      "learning_rate": 9.56951574290161e-06,
+      "loss": 0.6246,
+      "step": 2497
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.014207828651746,
+      "learning_rate": 9.569096891481979e-06,
+      "loss": 0.624,
+      "step": 2498
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5695149473103667,
+      "learning_rate": 9.568677845570497e-06,
+      "loss": 0.5503,
+      "step": 2499
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5961049589529291,
+      "learning_rate": 9.568258605184996e-06,
+      "loss": 0.4863,
+      "step": 2500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6857720938128696,
+      "learning_rate": 9.567839170343327e-06,
+      "loss": 0.5623,
+      "step": 2501
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8282032857138397,
+      "learning_rate": 9.567419541063338e-06,
+      "loss": 0.5572,
+      "step": 2502
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6893026749200307,
+      "learning_rate": 9.566999717362897e-06,
+      "loss": 0.5443,
+      "step": 2503
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7768239746749537,
+      "learning_rate": 9.56657969925987e-06,
+      "loss": 0.5686,
+      "step": 2504
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5597594429832988,
+      "learning_rate": 9.56615948677214e-06,
+      "loss": 0.5284,
+      "step": 2505
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7564168137120164,
+      "learning_rate": 9.565739079917591e-06,
+      "loss": 0.5774,
+      "step": 2506
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7641672679370525,
+      "learning_rate": 9.56531847871412e-06,
+      "loss": 0.6491,
+      "step": 2507
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9111016128120439,
+      "learning_rate": 9.56489768317963e-06,
+      "loss": 0.6339,
+      "step": 2508
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8567261134162234,
+      "learning_rate": 9.564476693332032e-06,
+      "loss": 0.6073,
+      "step": 2509
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8562793246556688,
+      "learning_rate": 9.564055509189247e-06,
+      "loss": 0.6335,
+      "step": 2510
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6186671603918202,
+      "learning_rate": 9.563634130769206e-06,
+      "loss": 0.5171,
+      "step": 2511
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.639540957253105,
+      "learning_rate": 9.56321255808984e-06,
+      "loss": 0.6051,
+      "step": 2512
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6801078881261133,
+      "learning_rate": 9.5627907911691e-06,
+      "loss": 0.5157,
+      "step": 2513
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6373131560688663,
+      "learning_rate": 9.562368830024935e-06,
+      "loss": 0.4953,
+      "step": 2514
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7733377559227137,
+      "learning_rate": 9.561946674675308e-06,
+      "loss": 0.6336,
+      "step": 2515
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8001748975946251,
+      "learning_rate": 9.561524325138192e-06,
+      "loss": 0.657,
+      "step": 2516
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8574515862304499,
+      "learning_rate": 9.561101781431558e-06,
+      "loss": 0.6496,
+      "step": 2517
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8432913593456502,
+      "learning_rate": 9.5606790435734e-06,
+      "loss": 0.6368,
+      "step": 2518
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7313829825713054,
+      "learning_rate": 9.560256111581705e-06,
+      "loss": 0.604,
+      "step": 2519
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6259317550178952,
+      "learning_rate": 9.559832985474482e-06,
+      "loss": 0.594,
+      "step": 2520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6732396527372642,
+      "learning_rate": 9.559409665269741e-06,
+      "loss": 0.5467,
+      "step": 2521
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9823587856496049,
+      "learning_rate": 9.5589861509855e-06,
+      "loss": 0.636,
+      "step": 2522
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6743790179134761,
+      "learning_rate": 9.558562442639785e-06,
+      "loss": 0.5974,
+      "step": 2523
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8780746135749471,
+      "learning_rate": 9.558138540250636e-06,
+      "loss": 0.6705,
+      "step": 2524
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8130945511228779,
+      "learning_rate": 9.557714443836093e-06,
+      "loss": 0.6641,
+      "step": 2525
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7906361142613465,
+      "learning_rate": 9.557290153414211e-06,
+      "loss": 0.636,
+      "step": 2526
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7101826950871989,
+      "learning_rate": 9.556865669003053e-06,
+      "loss": 0.5826,
+      "step": 2527
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6076279506595039,
+      "learning_rate": 9.556440990620682e-06,
+      "loss": 0.5123,
+      "step": 2528
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6120283623210143,
+      "learning_rate": 9.556016118285178e-06,
+      "loss": 0.5538,
+      "step": 2529
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7650111755183286,
+      "learning_rate": 9.555591052014629e-06,
+      "loss": 0.5438,
+      "step": 2530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5349207680266397,
+      "learning_rate": 9.555165791827125e-06,
+      "loss": 0.5127,
+      "step": 2531
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7083003864694378,
+      "learning_rate": 9.55474033774077e-06,
+      "loss": 0.5487,
+      "step": 2532
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.079547641280141,
+      "learning_rate": 9.554314689773674e-06,
+      "loss": 0.6343,
+      "step": 2533
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6465127203315632,
+      "learning_rate": 9.553888847943956e-06,
+      "loss": 0.5409,
+      "step": 2534
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6650303012598022,
+      "learning_rate": 9.55346281226974e-06,
+      "loss": 0.5893,
+      "step": 2535
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6110300011948008,
+      "learning_rate": 9.553036582769164e-06,
+      "loss": 0.4981,
+      "step": 2536
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8100895984954982,
+      "learning_rate": 9.552610159460369e-06,
+      "loss": 0.6186,
+      "step": 2537
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8179220677587817,
+      "learning_rate": 9.552183542361508e-06,
+      "loss": 0.5529,
+      "step": 2538
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5952520994167613,
+      "learning_rate": 9.551756731490742e-06,
+      "loss": 0.522,
+      "step": 2539
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8226784584113425,
+      "learning_rate": 9.551329726866235e-06,
+      "loss": 0.5815,
+      "step": 2540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5704498044176846,
+      "learning_rate": 9.550902528506168e-06,
+      "loss": 0.4416,
+      "step": 2541
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.604239828616303,
+      "learning_rate": 9.550475136428721e-06,
+      "loss": 0.5282,
+      "step": 2542
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.649238403244341,
+      "learning_rate": 9.550047550652089e-06,
+      "loss": 0.5562,
+      "step": 2543
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6570463366363107,
+      "learning_rate": 9.549619771194472e-06,
+      "loss": 0.5762,
+      "step": 2544
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6351487994969179,
+      "learning_rate": 9.54919179807408e-06,
+      "loss": 0.5349,
+      "step": 2545
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.648900949394083,
+      "learning_rate": 9.548763631309132e-06,
+      "loss": 0.5566,
+      "step": 2546
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5933008133983052,
+      "learning_rate": 9.54833527091785e-06,
+      "loss": 0.5216,
+      "step": 2547
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5900970115260598,
+      "learning_rate": 9.547906716918472e-06,
+      "loss": 0.4871,
+      "step": 2548
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6477875891595827,
+      "learning_rate": 9.547477969329235e-06,
+      "loss": 0.5577,
+      "step": 2549
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5294932860860582,
+      "learning_rate": 9.547049028168395e-06,
+      "loss": 0.4855,
+      "step": 2550
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7605284892461301,
+      "learning_rate": 9.546619893454208e-06,
+      "loss": 0.5526,
+      "step": 2551
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8821681184089893,
+      "learning_rate": 9.546190565204941e-06,
+      "loss": 0.6431,
+      "step": 2552
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5972036879598572,
+      "learning_rate": 9.545761043438868e-06,
+      "loss": 0.5477,
+      "step": 2553
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7774109891516569,
+      "learning_rate": 9.545331328174274e-06,
+      "loss": 0.6452,
+      "step": 2554
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.559993123168119,
+      "learning_rate": 9.544901419429452e-06,
+      "loss": 0.4755,
+      "step": 2555
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9812318418974344,
+      "learning_rate": 9.544471317222696e-06,
+      "loss": 0.631,
+      "step": 2556
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5700560989878682,
+      "learning_rate": 9.54404102157232e-06,
+      "loss": 0.5152,
+      "step": 2557
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6060204128545266,
+      "learning_rate": 9.54361053249664e-06,
+      "loss": 0.48,
+      "step": 2558
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.682067997060381,
+      "learning_rate": 9.543179850013978e-06,
+      "loss": 0.5346,
+      "step": 2559
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1084238206703774,
+      "learning_rate": 9.542748974142668e-06,
+      "loss": 0.6243,
+      "step": 2560
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8378037363051217,
+      "learning_rate": 9.542317904901049e-06,
+      "loss": 0.5972,
+      "step": 2561
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7135604533169227,
+      "learning_rate": 9.541886642307473e-06,
+      "loss": 0.6139,
+      "step": 2562
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6945523477457736,
+      "learning_rate": 9.541455186380297e-06,
+      "loss": 0.556,
+      "step": 2563
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8105865099203851,
+      "learning_rate": 9.541023537137885e-06,
+      "loss": 0.662,
+      "step": 2564
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9723262911446988,
+      "learning_rate": 9.540591694598615e-06,
+      "loss": 0.6237,
+      "step": 2565
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6284942090793638,
+      "learning_rate": 9.540159658780862e-06,
+      "loss": 0.5601,
+      "step": 2566
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7943621277394831,
+      "learning_rate": 9.539727429703024e-06,
+      "loss": 0.6296,
+      "step": 2567
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.646599463677551,
+      "learning_rate": 9.539295007383496e-06,
+      "loss": 0.5127,
+      "step": 2568
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.714891775660591,
+      "learning_rate": 9.538862391840683e-06,
+      "loss": 0.5886,
+      "step": 2569
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7683683299777343,
+      "learning_rate": 9.538429583093003e-06,
+      "loss": 0.6505,
+      "step": 2570
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8068531708286238,
+      "learning_rate": 9.537996581158878e-06,
+      "loss": 0.5425,
+      "step": 2571
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7633809149010669,
+      "learning_rate": 9.53756338605674e-06,
+      "loss": 0.5695,
+      "step": 2572
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5927862910594637,
+      "learning_rate": 9.53712999780503e-06,
+      "loss": 0.5093,
+      "step": 2573
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6329786480367147,
+      "learning_rate": 9.536696416422194e-06,
+      "loss": 0.533,
+      "step": 2574
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7403403535131283,
+      "learning_rate": 9.536262641926689e-06,
+      "loss": 0.6452,
+      "step": 2575
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5918314936786386,
+      "learning_rate": 9.535828674336976e-06,
+      "loss": 0.4818,
+      "step": 2576
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5683143465029358,
+      "learning_rate": 9.535394513671534e-06,
+      "loss": 0.4978,
+      "step": 2577
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.233205772739789,
+      "learning_rate": 9.534960159948841e-06,
+      "loss": 0.6469,
+      "step": 2578
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9695389770572457,
+      "learning_rate": 9.534525613187383e-06,
+      "loss": 0.6816,
+      "step": 2579
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5381105391785254,
+      "learning_rate": 9.534090873405662e-06,
+      "loss": 0.511,
+      "step": 2580
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.190502374810711,
+      "learning_rate": 9.53365594062218e-06,
+      "loss": 0.5892,
+      "step": 2581
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8059850644306538,
+      "learning_rate": 9.533220814855453e-06,
+      "loss": 0.6301,
+      "step": 2582
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6498408537331678,
+      "learning_rate": 9.532785496124003e-06,
+      "loss": 0.5565,
+      "step": 2583
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5783800531387244,
+      "learning_rate": 9.532349984446358e-06,
+      "loss": 0.5364,
+      "step": 2584
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9546698067137003,
+      "learning_rate": 9.531914279841057e-06,
+      "loss": 0.5862,
+      "step": 2585
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5701886528114443,
+      "learning_rate": 9.531478382326647e-06,
+      "loss": 0.5115,
+      "step": 2586
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6303662784078251,
+      "learning_rate": 9.531042291921685e-06,
+      "loss": 0.5237,
+      "step": 2587
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6637468226996036,
+      "learning_rate": 9.53060600864473e-06,
+      "loss": 0.4938,
+      "step": 2588
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7976804502105883,
+      "learning_rate": 9.530169532514355e-06,
+      "loss": 0.4904,
+      "step": 2589
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8480503149393777,
+      "learning_rate": 9.52973286354914e-06,
+      "loss": 0.6147,
+      "step": 2590
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5410025719543855,
+      "learning_rate": 9.529296001767669e-06,
+      "loss": 0.4789,
+      "step": 2591
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6944545405994262,
+      "learning_rate": 9.528858947188543e-06,
+      "loss": 0.5154,
+      "step": 2592
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5943387449210243,
+      "learning_rate": 9.528421699830365e-06,
+      "loss": 0.5479,
+      "step": 2593
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7460414739115181,
+      "learning_rate": 9.527984259711744e-06,
+      "loss": 0.5879,
+      "step": 2594
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6829949307460953,
+      "learning_rate": 9.527546626851306e-06,
+      "loss": 0.612,
+      "step": 2595
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7688717836898237,
+      "learning_rate": 9.527108801267673e-06,
+      "loss": 0.5747,
+      "step": 2596
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6935414862682998,
+      "learning_rate": 9.526670782979483e-06,
+      "loss": 0.5645,
+      "step": 2597
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.91662531566737,
+      "learning_rate": 9.526232572005387e-06,
+      "loss": 0.6257,
+      "step": 2598
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6404437087562576,
+      "learning_rate": 9.525794168364032e-06,
+      "loss": 0.5792,
+      "step": 2599
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9336752361852743,
+      "learning_rate": 9.525355572074083e-06,
+      "loss": 0.6525,
+      "step": 2600
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6188451823758374,
+      "learning_rate": 9.524916783154206e-06,
+      "loss": 0.5327,
+      "step": 2601
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.595947713758538,
+      "learning_rate": 9.524477801623083e-06,
+      "loss": 0.5018,
+      "step": 2602
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7602569644562919,
+      "learning_rate": 9.524038627499396e-06,
+      "loss": 0.5526,
+      "step": 2603
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1390855016538206,
+      "learning_rate": 9.523599260801845e-06,
+      "loss": 0.6233,
+      "step": 2604
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5826783634067593,
+      "learning_rate": 9.523159701549126e-06,
+      "loss": 0.4934,
+      "step": 2605
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6096115493108369,
+      "learning_rate": 9.522719949759953e-06,
+      "loss": 0.5076,
+      "step": 2606
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7390090601415761,
+      "learning_rate": 9.522280005453045e-06,
+      "loss": 0.6452,
+      "step": 2607
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6902664668253552,
+      "learning_rate": 9.521839868647128e-06,
+      "loss": 0.585,
+      "step": 2608
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8791964525039503,
+      "learning_rate": 9.521399539360937e-06,
+      "loss": 0.6836,
+      "step": 2609
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6490246657183921,
+      "learning_rate": 9.520959017613215e-06,
+      "loss": 0.5153,
+      "step": 2610
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5809035468751219,
+      "learning_rate": 9.520518303422718e-06,
+      "loss": 0.5684,
+      "step": 2611
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6707458810382166,
+      "learning_rate": 9.5200773968082e-06,
+      "loss": 0.5169,
+      "step": 2612
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6345208821443088,
+      "learning_rate": 9.51963629778843e-06,
+      "loss": 0.5915,
+      "step": 2613
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.530845563877577,
+      "learning_rate": 9.519195006382188e-06,
+      "loss": 0.517,
+      "step": 2614
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.835407233106553,
+      "learning_rate": 9.518753522608255e-06,
+      "loss": 0.6703,
+      "step": 2615
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6793471378124661,
+      "learning_rate": 9.518311846485424e-06,
+      "loss": 0.5751,
+      "step": 2616
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7349933974032022,
+      "learning_rate": 9.517869978032497e-06,
+      "loss": 0.5582,
+      "step": 2617
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.074984731694823,
+      "learning_rate": 9.51742791726828e-06,
+      "loss": 0.6236,
+      "step": 2618
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9201587831352932,
+      "learning_rate": 9.516985664211595e-06,
+      "loss": 0.5787,
+      "step": 2619
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6485537188339017,
+      "learning_rate": 9.516543218881264e-06,
+      "loss": 0.5675,
+      "step": 2620
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7191479145391688,
+      "learning_rate": 9.51610058129612e-06,
+      "loss": 0.5694,
+      "step": 2621
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8073994848174533,
+      "learning_rate": 9.515657751475005e-06,
+      "loss": 0.653,
+      "step": 2622
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.620470519151046,
+      "learning_rate": 9.51521472943677e-06,
+      "loss": 0.5594,
+      "step": 2623
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8577981010508564,
+      "learning_rate": 9.514771515200273e-06,
+      "loss": 0.5185,
+      "step": 2624
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5613299806463133,
+      "learning_rate": 9.51432810878438e-06,
+      "loss": 0.5347,
+      "step": 2625
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6905584167856317,
+      "learning_rate": 9.513884510207965e-06,
+      "loss": 0.5404,
+      "step": 2626
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5787359917483325,
+      "learning_rate": 9.51344071948991e-06,
+      "loss": 0.5663,
+      "step": 2627
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6451525074311069,
+      "learning_rate": 9.512996736649107e-06,
+      "loss": 0.5533,
+      "step": 2628
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6246696688029569,
+      "learning_rate": 9.512552561704454e-06,
+      "loss": 0.5496,
+      "step": 2629
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6815441549936629,
+      "learning_rate": 9.51210819467486e-06,
+      "loss": 0.5343,
+      "step": 2630
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8360184672663747,
+      "learning_rate": 9.511663635579237e-06,
+      "loss": 0.657,
+      "step": 2631
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7001008146479152,
+      "learning_rate": 9.51121888443651e-06,
+      "loss": 0.6144,
+      "step": 2632
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.832107455918356,
+      "learning_rate": 9.510773941265612e-06,
+      "loss": 0.5801,
+      "step": 2633
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6515346390702125,
+      "learning_rate": 9.510328806085483e-06,
+      "loss": 0.536,
+      "step": 2634
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5205168551152277,
+      "learning_rate": 9.509883478915068e-06,
+      "loss": 0.4978,
+      "step": 2635
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8758102289762094,
+      "learning_rate": 9.509437959773324e-06,
+      "loss": 0.6378,
+      "step": 2636
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7007231004619369,
+      "learning_rate": 9.508992248679217e-06,
+      "loss": 0.5801,
+      "step": 2637
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6204772699830503,
+      "learning_rate": 9.50854634565172e-06,
+      "loss": 0.6512,
+      "step": 2638
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.969055388010146,
+      "learning_rate": 9.508100250709809e-06,
+      "loss": 0.6135,
+      "step": 2639
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1653372937097857,
+      "learning_rate": 9.507653963872479e-06,
+      "loss": 0.5752,
+      "step": 2640
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6896592730666222,
+      "learning_rate": 9.507207485158724e-06,
+      "loss": 0.5448,
+      "step": 2641
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0668974797658888,
+      "learning_rate": 9.506760814587547e-06,
+      "loss": 0.6443,
+      "step": 2642
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8925707789010904,
+      "learning_rate": 9.506313952177966e-06,
+      "loss": 0.6913,
+      "step": 2643
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7922544781589685,
+      "learning_rate": 9.505866897949e-06,
+      "loss": 0.5695,
+      "step": 2644
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5977798601691642,
+      "learning_rate": 9.505419651919678e-06,
+      "loss": 0.5472,
+      "step": 2645
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5857657542808271,
+      "learning_rate": 9.50497221410904e-06,
+      "loss": 0.5256,
+      "step": 2646
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6269816345974752,
+      "learning_rate": 9.50452458453613e-06,
+      "loss": 0.5469,
+      "step": 2647
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6637418482259861,
+      "learning_rate": 9.504076763220003e-06,
+      "loss": 0.5697,
+      "step": 2648
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7637371397015333,
+      "learning_rate": 9.503628750179719e-06,
+      "loss": 0.568,
+      "step": 2649
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.939164956586822,
+      "learning_rate": 9.503180545434353e-06,
+      "loss": 0.6734,
+      "step": 2650
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8160952788765521,
+      "learning_rate": 9.50273214900298e-06,
+      "loss": 0.589,
+      "step": 2651
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.664853124608123,
+      "learning_rate": 9.50228356090469e-06,
+      "loss": 0.5112,
+      "step": 2652
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6596469254493981,
+      "learning_rate": 9.501834781158574e-06,
+      "loss": 0.5901,
+      "step": 2653
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6962722835328786,
+      "learning_rate": 9.501385809783739e-06,
+      "loss": 0.5398,
+      "step": 2654
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6632304354029682,
+      "learning_rate": 9.500936646799293e-06,
+      "loss": 0.5534,
+      "step": 2655
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.847071596337105,
+      "learning_rate": 9.500487292224357e-06,
+      "loss": 0.6602,
+      "step": 2656
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6594231855021783,
+      "learning_rate": 9.500037746078058e-06,
+      "loss": 0.5528,
+      "step": 2657
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6536904250392873,
+      "learning_rate": 9.499588008379534e-06,
+      "loss": 0.5198,
+      "step": 2658
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6360593846325148,
+      "learning_rate": 9.499138079147927e-06,
+      "loss": 0.592,
+      "step": 2659
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6326183225471013,
+      "learning_rate": 9.498687958402388e-06,
+      "loss": 0.5683,
+      "step": 2660
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6737489838915545,
+      "learning_rate": 9.498237646162078e-06,
+      "loss": 0.5475,
+      "step": 2661
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6694602163920368,
+      "learning_rate": 9.497787142446166e-06,
+      "loss": 0.5977,
+      "step": 2662
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6919396971952262,
+      "learning_rate": 9.497336447273828e-06,
+      "loss": 0.6119,
+      "step": 2663
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6545875727828754,
+      "learning_rate": 9.496885560664251e-06,
+      "loss": 0.5487,
+      "step": 2664
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8161631128077677,
+      "learning_rate": 9.496434482636623e-06,
+      "loss": 0.6141,
+      "step": 2665
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.066318647899695,
+      "learning_rate": 9.49598321321015e-06,
+      "loss": 0.5804,
+      "step": 2666
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6145505682433737,
+      "learning_rate": 9.495531752404036e-06,
+      "loss": 0.5522,
+      "step": 2667
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7404581570988357,
+      "learning_rate": 9.495080100237503e-06,
+      "loss": 0.6437,
+      "step": 2668
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6453979666377999,
+      "learning_rate": 9.494628256729774e-06,
+      "loss": 0.5338,
+      "step": 2669
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7390857664877688,
+      "learning_rate": 9.494176221900081e-06,
+      "loss": 0.5518,
+      "step": 2670
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6313898819588964,
+      "learning_rate": 9.49372399576767e-06,
+      "loss": 0.5381,
+      "step": 2671
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7259062340874922,
+      "learning_rate": 9.493271578351787e-06,
+      "loss": 0.5903,
+      "step": 2672
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9063405020904501,
+      "learning_rate": 9.492818969671693e-06,
+      "loss": 0.5912,
+      "step": 2673
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6202276898209159,
+      "learning_rate": 9.492366169746652e-06,
+      "loss": 0.5696,
+      "step": 2674
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8134286613739936,
+      "learning_rate": 9.491913178595937e-06,
+      "loss": 0.596,
+      "step": 2675
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8993303631152776,
+      "learning_rate": 9.491459996238834e-06,
+      "loss": 0.641,
+      "step": 2676
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.816233493485614,
+      "learning_rate": 9.49100662269463e-06,
+      "loss": 0.6091,
+      "step": 2677
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6127999309058696,
+      "learning_rate": 9.490553057982627e-06,
+      "loss": 0.4907,
+      "step": 2678
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6327981151154719,
+      "learning_rate": 9.490099302122129e-06,
+      "loss": 0.5736,
+      "step": 2679
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8263889646504642,
+      "learning_rate": 9.489645355132452e-06,
+      "loss": 0.6461,
+      "step": 2680
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6240105661629423,
+      "learning_rate": 9.489191217032919e-06,
+      "loss": 0.5742,
+      "step": 2681
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6009437124199746,
+      "learning_rate": 9.488736887842862e-06,
+      "loss": 0.5071,
+      "step": 2682
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8700586532006659,
+      "learning_rate": 9.488282367581619e-06,
+      "loss": 0.712,
+      "step": 2683
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7503216585904273,
+      "learning_rate": 9.487827656268539e-06,
+      "loss": 0.5549,
+      "step": 2684
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.860642292830429,
+      "learning_rate": 9.487372753922976e-06,
+      "loss": 0.64,
+      "step": 2685
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5906724271653871,
+      "learning_rate": 9.486917660564294e-06,
+      "loss": 0.5373,
+      "step": 2686
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.748581513417254,
+      "learning_rate": 9.486462376211866e-06,
+      "loss": 0.5543,
+      "step": 2687
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6625302637533109,
+      "learning_rate": 9.486006900885073e-06,
+      "loss": 0.5095,
+      "step": 2688
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6821843085540856,
+      "learning_rate": 9.485551234603299e-06,
+      "loss": 0.5771,
+      "step": 2689
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8304659259874845,
+      "learning_rate": 9.485095377385943e-06,
+      "loss": 0.5634,
+      "step": 2690
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7512685960757559,
+      "learning_rate": 9.484639329252409e-06,
+      "loss": 0.6491,
+      "step": 2691
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7534347072646292,
+      "learning_rate": 9.48418309022211e-06,
+      "loss": 0.6229,
+      "step": 2692
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6126461797913709,
+      "learning_rate": 9.483726660314464e-06,
+      "loss": 0.5082,
+      "step": 2693
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7279374182617983,
+      "learning_rate": 9.483270039548906e-06,
+      "loss": 0.5317,
+      "step": 2694
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9354531491803656,
+      "learning_rate": 9.482813227944868e-06,
+      "loss": 0.6301,
+      "step": 2695
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2255087844545716,
+      "learning_rate": 9.482356225521794e-06,
+      "loss": 0.6343,
+      "step": 2696
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6515432382342676,
+      "learning_rate": 9.48189903229914e-06,
+      "loss": 0.608,
+      "step": 2697
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7259746189084652,
+      "learning_rate": 9.481441648296367e-06,
+      "loss": 0.6028,
+      "step": 2698
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8509618283107478,
+      "learning_rate": 9.480984073532943e-06,
+      "loss": 0.6006,
+      "step": 2699
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.661686115769553,
+      "learning_rate": 9.480526308028345e-06,
+      "loss": 0.5693,
+      "step": 2700
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8340500821731865,
+      "learning_rate": 9.480068351802063e-06,
+      "loss": 0.6269,
+      "step": 2701
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7294376986355238,
+      "learning_rate": 9.479610204873586e-06,
+      "loss": 0.5847,
+      "step": 2702
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.64222982626784,
+      "learning_rate": 9.479151867262417e-06,
+      "loss": 0.6098,
+      "step": 2703
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7411094065018148,
+      "learning_rate": 9.478693338988065e-06,
+      "loss": 0.5866,
+      "step": 2704
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7757412492305917,
+      "learning_rate": 9.47823462007005e-06,
+      "loss": 0.6302,
+      "step": 2705
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7008346031950602,
+      "learning_rate": 9.477775710527898e-06,
+      "loss": 0.6065,
+      "step": 2706
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6450380535106405,
+      "learning_rate": 9.477316610381144e-06,
+      "loss": 0.5669,
+      "step": 2707
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8020350688024981,
+      "learning_rate": 9.47685731964933e-06,
+      "loss": 0.5942,
+      "step": 2708
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6515230738536907,
+      "learning_rate": 9.476397838352003e-06,
+      "loss": 0.537,
+      "step": 2709
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7106797304009778,
+      "learning_rate": 9.475938166508726e-06,
+      "loss": 0.5968,
+      "step": 2710
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6767112085036076,
+      "learning_rate": 9.475478304139064e-06,
+      "loss": 0.5551,
+      "step": 2711
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8966197085509974,
+      "learning_rate": 9.475018251262594e-06,
+      "loss": 0.6464,
+      "step": 2712
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8575931820709087,
+      "learning_rate": 9.474558007898895e-06,
+      "loss": 0.6516,
+      "step": 2713
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9080998604636245,
+      "learning_rate": 9.47409757406756e-06,
+      "loss": 0.604,
+      "step": 2714
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8304634081485033,
+      "learning_rate": 9.47363694978819e-06,
+      "loss": 0.5593,
+      "step": 2715
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8461238243280341,
+      "learning_rate": 9.473176135080392e-06,
+      "loss": 0.6296,
+      "step": 2716
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6635571220376301,
+      "learning_rate": 9.472715129963776e-06,
+      "loss": 0.5634,
+      "step": 2717
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6624474616243881,
+      "learning_rate": 9.472253934457973e-06,
+      "loss": 0.5372,
+      "step": 2718
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6674988841873706,
+      "learning_rate": 9.47179254858261e-06,
+      "loss": 0.5388,
+      "step": 2719
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.19573730107884,
+      "learning_rate": 9.471330972357327e-06,
+      "loss": 0.6058,
+      "step": 2720
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5937215067510677,
+      "learning_rate": 9.470869205801774e-06,
+      "loss": 0.566,
+      "step": 2721
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5487173465229642,
+      "learning_rate": 9.470407248935606e-06,
+      "loss": 0.5182,
+      "step": 2722
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8008315113990622,
+      "learning_rate": 9.469945101778487e-06,
+      "loss": 0.5512,
+      "step": 2723
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6749228954537383,
+      "learning_rate": 9.46948276435009e-06,
+      "loss": 0.5427,
+      "step": 2724
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7251223243488515,
+      "learning_rate": 9.46902023667009e-06,
+      "loss": 0.5931,
+      "step": 2725
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5893424590220128,
+      "learning_rate": 9.468557518758183e-06,
+      "loss": 0.5271,
+      "step": 2726
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6206131778648086,
+      "learning_rate": 9.468094610634062e-06,
+      "loss": 0.5336,
+      "step": 2727
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7232629802065703,
+      "learning_rate": 9.46763151231743e-06,
+      "loss": 0.5801,
+      "step": 2728
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5559640325030679,
+      "learning_rate": 9.467168223828001e-06,
+      "loss": 0.5295,
+      "step": 2729
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6453014003908335,
+      "learning_rate": 9.466704745185497e-06,
+      "loss": 0.5208,
+      "step": 2730
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7949818119663074,
+      "learning_rate": 9.466241076409644e-06,
+      "loss": 0.6,
+      "step": 2731
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6768070524933107,
+      "learning_rate": 9.46577721752018e-06,
+      "loss": 0.5493,
+      "step": 2732
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6235733846150524,
+      "learning_rate": 9.465313168536852e-06,
+      "loss": 0.4976,
+      "step": 2733
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5526313657564148,
+      "learning_rate": 9.464848929479412e-06,
+      "loss": 0.5021,
+      "step": 2734
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7255657737830097,
+      "learning_rate": 9.46438450036762e-06,
+      "loss": 0.5794,
+      "step": 2735
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.729547906615159,
+      "learning_rate": 9.463919881221246e-06,
+      "loss": 0.631,
+      "step": 2736
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7358070039119499,
+      "learning_rate": 9.463455072060066e-06,
+      "loss": 0.7151,
+      "step": 2737
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.565924718889971,
+      "learning_rate": 9.462990072903868e-06,
+      "loss": 0.4992,
+      "step": 2738
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.708636311457162,
+      "learning_rate": 9.462524883772444e-06,
+      "loss": 0.5494,
+      "step": 2739
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7565394219576851,
+      "learning_rate": 9.462059504685596e-06,
+      "loss": 0.6017,
+      "step": 2740
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6288142162213577,
+      "learning_rate": 9.461593935663134e-06,
+      "loss": 0.5263,
+      "step": 2741
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.628364933812235,
+      "learning_rate": 9.461128176724876e-06,
+      "loss": 0.522,
+      "step": 2742
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6109788086616127,
+      "learning_rate": 9.460662227890645e-06,
+      "loss": 0.55,
+      "step": 2743
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.710527864029022,
+      "learning_rate": 9.460196089180278e-06,
+      "loss": 0.6023,
+      "step": 2744
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5793125193996272,
+      "learning_rate": 9.459729760613618e-06,
+      "loss": 0.521,
+      "step": 2745
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.633340780233596,
+      "learning_rate": 9.459263242210511e-06,
+      "loss": 0.5724,
+      "step": 2746
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9481665252210578,
+      "learning_rate": 9.45879653399082e-06,
+      "loss": 0.6289,
+      "step": 2747
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7203959862754759,
+      "learning_rate": 9.458329635974405e-06,
+      "loss": 0.6071,
+      "step": 2748
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7140244824243365,
+      "learning_rate": 9.457862548181147e-06,
+      "loss": 0.5492,
+      "step": 2749
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6796579025760305,
+      "learning_rate": 9.457395270630926e-06,
+      "loss": 0.5862,
+      "step": 2750
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6233384318469098,
+      "learning_rate": 9.456927803343631e-06,
+      "loss": 0.6139,
+      "step": 2751
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5590310560221577,
+      "learning_rate": 9.456460146339162e-06,
+      "loss": 0.5194,
+      "step": 2752
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6587705455020467,
+      "learning_rate": 9.455992299637426e-06,
+      "loss": 0.5528,
+      "step": 2753
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6777434539217504,
+      "learning_rate": 9.455524263258335e-06,
+      "loss": 0.5015,
+      "step": 2754
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.218106107899709,
+      "learning_rate": 9.455056037221816e-06,
+      "loss": 0.5961,
+      "step": 2755
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8525734012370407,
+      "learning_rate": 9.454587621547796e-06,
+      "loss": 0.6683,
+      "step": 2756
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9978909741872638,
+      "learning_rate": 9.454119016256218e-06,
+      "loss": 0.6548,
+      "step": 2757
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6537367828334337,
+      "learning_rate": 9.453650221367025e-06,
+      "loss": 0.5757,
+      "step": 2758
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8057521803778169,
+      "learning_rate": 9.453181236900175e-06,
+      "loss": 0.5573,
+      "step": 2759
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8751902451285365,
+      "learning_rate": 9.452712062875631e-06,
+      "loss": 0.603,
+      "step": 2760
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7558656559155569,
+      "learning_rate": 9.452242699313361e-06,
+      "loss": 0.6003,
+      "step": 2761
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.643190086717265,
+      "learning_rate": 9.451773146233348e-06,
+      "loss": 0.5123,
+      "step": 2762
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7753792676439469,
+      "learning_rate": 9.451303403655579e-06,
+      "loss": 0.6313,
+      "step": 2763
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7982668840577974,
+      "learning_rate": 9.450833471600047e-06,
+      "loss": 0.6137,
+      "step": 2764
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6469800456142494,
+      "learning_rate": 9.450363350086756e-06,
+      "loss": 0.5721,
+      "step": 2765
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5954827834637731,
+      "learning_rate": 9.44989303913572e-06,
+      "loss": 0.5192,
+      "step": 2766
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6267961389988276,
+      "learning_rate": 9.449422538766958e-06,
+      "loss": 0.5352,
+      "step": 2767
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6181168997578211,
+      "learning_rate": 9.448951849000494e-06,
+      "loss": 0.5943,
+      "step": 2768
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9379775552391613,
+      "learning_rate": 9.448480969856368e-06,
+      "loss": 0.6026,
+      "step": 2769
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8224891932385208,
+      "learning_rate": 9.448009901354623e-06,
+      "loss": 0.6197,
+      "step": 2770
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6413797970707639,
+      "learning_rate": 9.44753864351531e-06,
+      "loss": 0.5587,
+      "step": 2771
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9317899250101933,
+      "learning_rate": 9.44706719635849e-06,
+      "loss": 0.6364,
+      "step": 2772
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6969392482600794,
+      "learning_rate": 9.446595559904228e-06,
+      "loss": 0.5118,
+      "step": 2773
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7753036998085789,
+      "learning_rate": 9.446123734172606e-06,
+      "loss": 0.5955,
+      "step": 2774
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1130667261540452,
+      "learning_rate": 9.445651719183701e-06,
+      "loss": 0.7081,
+      "step": 2775
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7918291910362325,
+      "learning_rate": 9.44517951495761e-06,
+      "loss": 0.624,
+      "step": 2776
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6756623501720954,
+      "learning_rate": 9.444707121514432e-06,
+      "loss": 0.5837,
+      "step": 2777
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6287568219254337,
+      "learning_rate": 9.444234538874273e-06,
+      "loss": 0.5435,
+      "step": 2778
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5790455136469947,
+      "learning_rate": 9.443761767057253e-06,
+      "loss": 0.517,
+      "step": 2779
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8612660254985647,
+      "learning_rate": 9.443288806083496e-06,
+      "loss": 0.6321,
+      "step": 2780
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7449823418496511,
+      "learning_rate": 9.442815655973133e-06,
+      "loss": 0.6455,
+      "step": 2781
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5845316912298215,
+      "learning_rate": 9.442342316746303e-06,
+      "loss": 0.555,
+      "step": 2782
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9020077596190232,
+      "learning_rate": 9.441868788423156e-06,
+      "loss": 0.6682,
+      "step": 2783
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6004706591343363,
+      "learning_rate": 9.44139507102385e-06,
+      "loss": 0.5382,
+      "step": 2784
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7774042995780771,
+      "learning_rate": 9.440921164568548e-06,
+      "loss": 0.6594,
+      "step": 2785
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7090563768533579,
+      "learning_rate": 9.440447069077425e-06,
+      "loss": 0.5637,
+      "step": 2786
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6212625374063201,
+      "learning_rate": 9.439972784570659e-06,
+      "loss": 0.5373,
+      "step": 2787
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9759312113547524,
+      "learning_rate": 9.439498311068438e-06,
+      "loss": 0.6406,
+      "step": 2788
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7079321917169328,
+      "learning_rate": 9.439023648590961e-06,
+      "loss": 0.5166,
+      "step": 2789
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6498521179541045,
+      "learning_rate": 9.438548797158435e-06,
+      "loss": 0.532,
+      "step": 2790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5737016888241293,
+      "learning_rate": 9.438073756791068e-06,
+      "loss": 0.5157,
+      "step": 2791
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6755265566833282,
+      "learning_rate": 9.437598527509082e-06,
+      "loss": 0.5655,
+      "step": 2792
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7078292393750686,
+      "learning_rate": 9.43712310933271e-06,
+      "loss": 0.6258,
+      "step": 2793
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7853443316626899,
+      "learning_rate": 9.436647502282185e-06,
+      "loss": 0.5843,
+      "step": 2794
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7130818120694572,
+      "learning_rate": 9.436171706377753e-06,
+      "loss": 0.5606,
+      "step": 2795
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.77522789020357,
+      "learning_rate": 9.435695721639668e-06,
+      "loss": 0.6855,
+      "step": 2796
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6393191307337324,
+      "learning_rate": 9.43521954808819e-06,
+      "loss": 0.5047,
+      "step": 2797
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8133557704088161,
+      "learning_rate": 9.43474318574359e-06,
+      "loss": 0.6843,
+      "step": 2798
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.91311500568812,
+      "learning_rate": 9.434266634626143e-06,
+      "loss": 0.6161,
+      "step": 2799
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6518119783303293,
+      "learning_rate": 9.433789894756136e-06,
+      "loss": 0.5574,
+      "step": 2800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2686676296534765,
+      "learning_rate": 9.43331296615386e-06,
+      "loss": 0.644,
+      "step": 2801
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.748624786857783,
+      "learning_rate": 9.432835848839619e-06,
+      "loss": 0.5264,
+      "step": 2802
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.638531267290233,
+      "learning_rate": 9.432358542833722e-06,
+      "loss": 0.5296,
+      "step": 2803
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6667117779795751,
+      "learning_rate": 9.431881048156484e-06,
+      "loss": 0.5383,
+      "step": 2804
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1193428143495239,
+      "learning_rate": 9.431403364828233e-06,
+      "loss": 0.5939,
+      "step": 2805
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.919471429477277,
+      "learning_rate": 9.4309254928693e-06,
+      "loss": 0.622,
+      "step": 2806
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.817986507762727,
+      "learning_rate": 9.43044743230003e-06,
+      "loss": 0.6453,
+      "step": 2807
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7075351440730898,
+      "learning_rate": 9.429969183140771e-06,
+      "loss": 0.5307,
+      "step": 2808
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8892681978788348,
+      "learning_rate": 9.429490745411878e-06,
+      "loss": 0.5703,
+      "step": 2809
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6934844745237143,
+      "learning_rate": 9.42901211913372e-06,
+      "loss": 0.5847,
+      "step": 2810
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7907377083152881,
+      "learning_rate": 9.428533304326668e-06,
+      "loss": 0.5943,
+      "step": 2811
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7895661295094039,
+      "learning_rate": 9.428054301011104e-06,
+      "loss": 0.6229,
+      "step": 2812
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.592793376722089,
+      "learning_rate": 9.42757510920742e-06,
+      "loss": 0.5125,
+      "step": 2813
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9090321305209546,
+      "learning_rate": 9.427095728936013e-06,
+      "loss": 0.6113,
+      "step": 2814
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6829019996273853,
+      "learning_rate": 9.426616160217288e-06,
+      "loss": 0.5814,
+      "step": 2815
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6471304331268625,
+      "learning_rate": 9.426136403071656e-06,
+      "loss": 0.5511,
+      "step": 2816
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6523135127801731,
+      "learning_rate": 9.425656457519544e-06,
+      "loss": 0.5825,
+      "step": 2817
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.159718982308664,
+      "learning_rate": 9.425176323581375e-06,
+      "loss": 0.6203,
+      "step": 2818
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8029568291454016,
+      "learning_rate": 9.424696001277594e-06,
+      "loss": 0.5665,
+      "step": 2819
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0670005911598865,
+      "learning_rate": 9.424215490628644e-06,
+      "loss": 0.6193,
+      "step": 2820
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6666859487893871,
+      "learning_rate": 9.423734791654976e-06,
+      "loss": 0.5869,
+      "step": 2821
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7380117096550212,
+      "learning_rate": 9.423253904377054e-06,
+      "loss": 0.5548,
+      "step": 2822
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.86011089010381,
+      "learning_rate": 9.42277282881535e-06,
+      "loss": 0.5972,
+      "step": 2823
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8826088904629107,
+      "learning_rate": 9.422291564990339e-06,
+      "loss": 0.6252,
+      "step": 2824
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8408935153832707,
+      "learning_rate": 9.421810112922507e-06,
+      "loss": 0.6092,
+      "step": 2825
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8537020795261052,
+      "learning_rate": 9.421328472632349e-06,
+      "loss": 0.68,
+      "step": 2826
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6949599215545131,
+      "learning_rate": 9.420846644140368e-06,
+      "loss": 0.5896,
+      "step": 2827
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7244250735563492,
+      "learning_rate": 9.420364627467071e-06,
+      "loss": 0.6225,
+      "step": 2828
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7446010533844636,
+      "learning_rate": 9.419882422632978e-06,
+      "loss": 0.588,
+      "step": 2829
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7872628454553385,
+      "learning_rate": 9.419400029658613e-06,
+      "loss": 0.5979,
+      "step": 2830
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6342756540961432,
+      "learning_rate": 9.418917448564512e-06,
+      "loss": 0.5236,
+      "step": 2831
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9368501245531178,
+      "learning_rate": 9.418434679371216e-06,
+      "loss": 0.6125,
+      "step": 2832
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.033354267299191,
+      "learning_rate": 9.417951722099275e-06,
+      "loss": 0.6481,
+      "step": 2833
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.630607770788086,
+      "learning_rate": 9.417468576769247e-06,
+      "loss": 0.5658,
+      "step": 2834
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4325380942899781,
+      "learning_rate": 9.416985243401696e-06,
+      "loss": 0.6096,
+      "step": 2835
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5761440335248742,
+      "learning_rate": 9.4165017220172e-06,
+      "loss": 0.5246,
+      "step": 2836
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7249543068030742,
+      "learning_rate": 9.41601801263634e-06,
+      "loss": 0.6412,
+      "step": 2837
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9412920957801415,
+      "learning_rate": 9.415534115279701e-06,
+      "loss": 0.5744,
+      "step": 2838
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7897584709082406,
+      "learning_rate": 9.415050029967887e-06,
+      "loss": 0.5695,
+      "step": 2839
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0703647721946201,
+      "learning_rate": 9.4145657567215e-06,
+      "loss": 0.595,
+      "step": 2840
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8092186795924167,
+      "learning_rate": 9.414081295561157e-06,
+      "loss": 0.6585,
+      "step": 2841
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9069107351169117,
+      "learning_rate": 9.41359664650748e-06,
+      "loss": 0.639,
+      "step": 2842
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7338329136891169,
+      "learning_rate": 9.413111809581097e-06,
+      "loss": 0.5871,
+      "step": 2843
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.645351721082941,
+      "learning_rate": 9.412626784802646e-06,
+      "loss": 0.5864,
+      "step": 2844
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.853078731366598,
+      "learning_rate": 9.412141572192772e-06,
+      "loss": 0.6837,
+      "step": 2845
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5737539198299346,
+      "learning_rate": 9.411656171772132e-06,
+      "loss": 0.5264,
+      "step": 2846
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8672541381913115,
+      "learning_rate": 9.411170583561386e-06,
+      "loss": 0.6583,
+      "step": 2847
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8309777279979316,
+      "learning_rate": 9.410684807581204e-06,
+      "loss": 0.6165,
+      "step": 2848
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5985005098021324,
+      "learning_rate": 9.410198843852267e-06,
+      "loss": 0.5366,
+      "step": 2849
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9759883360897562,
+      "learning_rate": 9.409712692395257e-06,
+      "loss": 0.6922,
+      "step": 2850
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0933014987638408,
+      "learning_rate": 9.409226353230866e-06,
+      "loss": 0.6349,
+      "step": 2851
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6851806530435292,
+      "learning_rate": 9.408739826379802e-06,
+      "loss": 0.5324,
+      "step": 2852
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6426987004506657,
+      "learning_rate": 9.40825311186277e-06,
+      "loss": 0.5423,
+      "step": 2853
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7543805886760679,
+      "learning_rate": 9.407766209700493e-06,
+      "loss": 0.6085,
+      "step": 2854
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6945186168163094,
+      "learning_rate": 9.40727911991369e-06,
+      "loss": 0.6244,
+      "step": 2855
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9530437827842488,
+      "learning_rate": 9.406791842523101e-06,
+      "loss": 0.6679,
+      "step": 2856
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6025847100072658,
+      "learning_rate": 9.406304377549464e-06,
+      "loss": 0.52,
+      "step": 2857
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8686059097275766,
+      "learning_rate": 9.40581672501353e-06,
+      "loss": 0.6455,
+      "step": 2858
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6122480572803359,
+      "learning_rate": 9.405328884936058e-06,
+      "loss": 0.5348,
+      "step": 2859
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.585461852705897,
+      "learning_rate": 9.404840857337814e-06,
+      "loss": 0.5403,
+      "step": 2860
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5677296145455577,
+      "learning_rate": 9.404352642239569e-06,
+      "loss": 0.5262,
+      "step": 2861
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.081751345125273,
+      "learning_rate": 9.403864239662106e-06,
+      "loss": 0.6406,
+      "step": 2862
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8885644008018558,
+      "learning_rate": 9.403375649626214e-06,
+      "loss": 0.6475,
+      "step": 2863
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6353144235774352,
+      "learning_rate": 9.402886872152695e-06,
+      "loss": 0.5235,
+      "step": 2864
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7409406000262078,
+      "learning_rate": 9.402397907262348e-06,
+      "loss": 0.6313,
+      "step": 2865
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5830889351646802,
+      "learning_rate": 9.401908754975993e-06,
+      "loss": 0.5451,
+      "step": 2866
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.601798767248453,
+      "learning_rate": 9.401419415314447e-06,
+      "loss": 0.501,
+      "step": 2867
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6403338198548815,
+      "learning_rate": 9.400929888298542e-06,
+      "loss": 0.583,
+      "step": 2868
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6519778667822522,
+      "learning_rate": 9.400440173949115e-06,
+      "loss": 0.5455,
+      "step": 2869
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.616829522867513,
+      "learning_rate": 9.399950272287011e-06,
+      "loss": 0.6113,
+      "step": 2870
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5984801817276588,
+      "learning_rate": 9.399460183333084e-06,
+      "loss": 0.4937,
+      "step": 2871
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8405439718925305,
+      "learning_rate": 9.398969907108198e-06,
+      "loss": 0.6504,
+      "step": 2872
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8285620852945064,
+      "learning_rate": 9.398479443633217e-06,
+      "loss": 0.5455,
+      "step": 2873
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.658413194806729,
+      "learning_rate": 9.397988792929024e-06,
+      "loss": 0.5504,
+      "step": 2874
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.779401416885058,
+      "learning_rate": 9.397497955016502e-06,
+      "loss": 0.5915,
+      "step": 2875
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5954922739090105,
+      "learning_rate": 9.397006929916542e-06,
+      "loss": 0.5742,
+      "step": 2876
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7908914462166622,
+      "learning_rate": 9.39651571765005e-06,
+      "loss": 0.6958,
+      "step": 2877
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5842953799188918,
+      "learning_rate": 9.396024318237932e-06,
+      "loss": 0.5165,
+      "step": 2878
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8778519407334405,
+      "learning_rate": 9.395532731701106e-06,
+      "loss": 0.6618,
+      "step": 2879
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.732685740487691,
+      "learning_rate": 9.3950409580605e-06,
+      "loss": 0.5762,
+      "step": 2880
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6341810910406681,
+      "learning_rate": 9.394548997337044e-06,
+      "loss": 0.5098,
+      "step": 2881
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6446350693674776,
+      "learning_rate": 9.394056849551681e-06,
+      "loss": 0.5658,
+      "step": 2882
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6860208191364687,
+      "learning_rate": 9.393564514725357e-06,
+      "loss": 0.5522,
+      "step": 2883
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.573984716627783,
+      "learning_rate": 9.393071992879033e-06,
+      "loss": 0.5381,
+      "step": 2884
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5983887150093357,
+      "learning_rate": 9.392579284033672e-06,
+      "loss": 0.5736,
+      "step": 2885
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7597083521338649,
+      "learning_rate": 9.392086388210248e-06,
+      "loss": 0.6242,
+      "step": 2886
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7071420000098009,
+      "learning_rate": 9.391593305429741e-06,
+      "loss": 0.5656,
+      "step": 2887
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7808994541893899,
+      "learning_rate": 9.39110003571314e-06,
+      "loss": 0.6728,
+      "step": 2888
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5811025956057759,
+      "learning_rate": 9.390606579081445e-06,
+      "loss": 0.5341,
+      "step": 2889
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6992273447442522,
+      "learning_rate": 9.390112935555655e-06,
+      "loss": 0.577,
+      "step": 2890
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7456521083506628,
+      "learning_rate": 9.389619105156788e-06,
+      "loss": 0.6125,
+      "step": 2891
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7156156740151897,
+      "learning_rate": 9.389125087905862e-06,
+      "loss": 0.6062,
+      "step": 2892
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.884844055039022,
+      "learning_rate": 9.388630883823909e-06,
+      "loss": 0.5962,
+      "step": 2893
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1687731001705837,
+      "learning_rate": 9.38813649293196e-06,
+      "loss": 0.6283,
+      "step": 2894
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5924002278826536,
+      "learning_rate": 9.387641915251065e-06,
+      "loss": 0.5487,
+      "step": 2895
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5738750677465778,
+      "learning_rate": 9.387147150802274e-06,
+      "loss": 0.589,
+      "step": 2896
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7334098818311762,
+      "learning_rate": 9.386652199606648e-06,
+      "loss": 0.5495,
+      "step": 2897
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6344325183000025,
+      "learning_rate": 9.386157061685255e-06,
+      "loss": 0.5708,
+      "step": 2898
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7188129179715578,
+      "learning_rate": 9.385661737059172e-06,
+      "loss": 0.5859,
+      "step": 2899
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5763663100946099,
+      "learning_rate": 9.385166225749485e-06,
+      "loss": 0.5492,
+      "step": 2900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7245777365379958,
+      "learning_rate": 9.384670527777284e-06,
+      "loss": 0.517,
+      "step": 2901
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8367286295807743,
+      "learning_rate": 9.38417464316367e-06,
+      "loss": 0.6508,
+      "step": 2902
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6331625799793746,
+      "learning_rate": 9.38367857192975e-06,
+      "loss": 0.5626,
+      "step": 2903
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8974882912021842,
+      "learning_rate": 9.383182314096643e-06,
+      "loss": 0.6413,
+      "step": 2904
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.715344356395881,
+      "learning_rate": 9.382685869685473e-06,
+      "loss": 0.5415,
+      "step": 2905
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7184999660613931,
+      "learning_rate": 9.382189238717367e-06,
+      "loss": 0.5578,
+      "step": 2906
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6660457606478688,
+      "learning_rate": 9.38169242121347e-06,
+      "loss": 0.5613,
+      "step": 2907
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7625046137579277,
+      "learning_rate": 9.38119541719493e-06,
+      "loss": 0.6153,
+      "step": 2908
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6219015803012954,
+      "learning_rate": 9.3806982266829e-06,
+      "loss": 0.5723,
+      "step": 2909
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6358396301577278,
+      "learning_rate": 9.380200849698547e-06,
+      "loss": 0.5039,
+      "step": 2910
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8309198004924665,
+      "learning_rate": 9.379703286263037e-06,
+      "loss": 0.6091,
+      "step": 2911
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7881318920771448,
+      "learning_rate": 9.379205536397558e-06,
+      "loss": 0.6043,
+      "step": 2912
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.658074606345979,
+      "learning_rate": 9.378707600123292e-06,
+      "loss": 0.5866,
+      "step": 2913
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7781760241293016,
+      "learning_rate": 9.378209477461435e-06,
+      "loss": 0.6365,
+      "step": 2914
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8019528651249476,
+      "learning_rate": 9.37771116843319e-06,
+      "loss": 0.6024,
+      "step": 2915
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6137475171688824,
+      "learning_rate": 9.377212673059771e-06,
+      "loss": 0.5373,
+      "step": 2916
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6083863516564999,
+      "learning_rate": 9.376713991362396e-06,
+      "loss": 0.5458,
+      "step": 2917
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8613242340804598,
+      "learning_rate": 9.376215123362293e-06,
+      "loss": 0.6441,
+      "step": 2918
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6604837134858038,
+      "learning_rate": 9.375716069080696e-06,
+      "loss": 0.5517,
+      "step": 2919
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7052519738588503,
+      "learning_rate": 9.375216828538849e-06,
+      "loss": 0.5298,
+      "step": 2920
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9483927267239565,
+      "learning_rate": 9.374717401758001e-06,
+      "loss": 0.6746,
+      "step": 2921
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.648908303677272,
+      "learning_rate": 9.374217788759417e-06,
+      "loss": 0.5798,
+      "step": 2922
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6520297891395666,
+      "learning_rate": 9.373717989564357e-06,
+      "loss": 0.5101,
+      "step": 2923
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6209113678896905,
+      "learning_rate": 9.373218004194098e-06,
+      "loss": 0.5482,
+      "step": 2924
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6624328566805205,
+      "learning_rate": 9.372717832669924e-06,
+      "loss": 0.6084,
+      "step": 2925
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6039012435279127,
+      "learning_rate": 9.372217475013125e-06,
+      "loss": 0.5126,
+      "step": 2926
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6613123256570572,
+      "learning_rate": 9.371716931245001e-06,
+      "loss": 0.576,
+      "step": 2927
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6287042674642637,
+      "learning_rate": 9.371216201386855e-06,
+      "loss": 0.5232,
+      "step": 2928
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5753287882600551,
+      "learning_rate": 9.370715285460006e-06,
+      "loss": 0.5522,
+      "step": 2929
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.846231266735744,
+      "learning_rate": 9.370214183485773e-06,
+      "loss": 0.6236,
+      "step": 2930
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7573771445645239,
+      "learning_rate": 9.369712895485487e-06,
+      "loss": 0.654,
+      "step": 2931
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.170271331259036,
+      "learning_rate": 9.369211421480488e-06,
+      "loss": 0.6071,
+      "step": 2932
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6709961556183903,
+      "learning_rate": 9.368709761492118e-06,
+      "loss": 0.5738,
+      "step": 2933
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6046082867343111,
+      "learning_rate": 9.368207915541736e-06,
+      "loss": 0.5495,
+      "step": 2934
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1325040102684705,
+      "learning_rate": 9.367705883650702e-06,
+      "loss": 0.6025,
+      "step": 2935
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7292815796694089,
+      "learning_rate": 9.367203665840385e-06,
+      "loss": 0.5687,
+      "step": 2936
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9575243393849735,
+      "learning_rate": 9.366701262132164e-06,
+      "loss": 0.6827,
+      "step": 2937
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6305451456666317,
+      "learning_rate": 9.366198672547424e-06,
+      "loss": 0.5839,
+      "step": 2938
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8588962449405341,
+      "learning_rate": 9.36569589710756e-06,
+      "loss": 0.6266,
+      "step": 2939
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.589765890407199,
+      "learning_rate": 9.365192935833972e-06,
+      "loss": 0.5277,
+      "step": 2940
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7664631674614872,
+      "learning_rate": 9.364689788748068e-06,
+      "loss": 0.6369,
+      "step": 2941
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1335936449339294,
+      "learning_rate": 9.36418645587127e-06,
+      "loss": 0.7047,
+      "step": 2942
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8685996226447723,
+      "learning_rate": 9.363682937224997e-06,
+      "loss": 0.6507,
+      "step": 2943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7300136339693422,
+      "learning_rate": 9.363179232830688e-06,
+      "loss": 0.5687,
+      "step": 2944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6508117148397068,
+      "learning_rate": 9.362675342709782e-06,
+      "loss": 0.5304,
+      "step": 2945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5731058920579397,
+      "learning_rate": 9.362171266883728e-06,
+      "loss": 0.5281,
+      "step": 2946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6330248233737237,
+      "learning_rate": 9.36166700537398e-06,
+      "loss": 0.6018,
+      "step": 2947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6931518987762789,
+      "learning_rate": 9.361162558202009e-06,
+      "loss": 0.5297,
+      "step": 2948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8090933159926346,
+      "learning_rate": 9.360657925389283e-06,
+      "loss": 0.6245,
+      "step": 2949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9081573626819215,
+      "learning_rate": 9.360153106957283e-06,
+      "loss": 0.6553,
+      "step": 2950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6089575980777817,
+      "learning_rate": 9.359648102927499e-06,
+      "loss": 0.5723,
+      "step": 2951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8969338297373163,
+      "learning_rate": 9.359142913321427e-06,
+      "loss": 0.6536,
+      "step": 2952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6081021581080863,
+      "learning_rate": 9.358637538160569e-06,
+      "loss": 0.5302,
+      "step": 2953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5845986707871152,
+      "learning_rate": 9.35813197746644e-06,
+      "loss": 0.5377,
+      "step": 2954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7618249368956932,
+      "learning_rate": 9.357626231260562e-06,
+      "loss": 0.5494,
+      "step": 2955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7053833506407878,
+      "learning_rate": 9.357120299564457e-06,
+      "loss": 0.5632,
+      "step": 2956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7454397296241266,
+      "learning_rate": 9.356614182399666e-06,
+      "loss": 0.5882,
+      "step": 2957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6079229664082009,
+      "learning_rate": 9.356107879787731e-06,
+      "loss": 0.5482,
+      "step": 2958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8058099202927756,
+      "learning_rate": 9.355601391750202e-06,
+      "loss": 0.5491,
+      "step": 2959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8408124520543254,
+      "learning_rate": 9.355094718308642e-06,
+      "loss": 0.7278,
+      "step": 2960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8638475470427798,
+      "learning_rate": 9.354587859484616e-06,
+      "loss": 0.6491,
+      "step": 2961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9616312733333742,
+      "learning_rate": 9.3540808152997e-06,
+      "loss": 0.6151,
+      "step": 2962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9157548569921015,
+      "learning_rate": 9.353573585775478e-06,
+      "loss": 0.6686,
+      "step": 2963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7716073816140209,
+      "learning_rate": 9.35306617093354e-06,
+      "loss": 0.616,
+      "step": 2964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7078847840826726,
+      "learning_rate": 9.352558570795485e-06,
+      "loss": 0.5714,
+      "step": 2965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6285416812475433,
+      "learning_rate": 9.35205078538292e-06,
+      "loss": 0.5332,
+      "step": 2966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.738607843354742,
+      "learning_rate": 9.351542814717463e-06,
+      "loss": 0.6244,
+      "step": 2967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7025995652133761,
+      "learning_rate": 9.351034658820731e-06,
+      "loss": 0.5678,
+      "step": 2968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6454152210839516,
+      "learning_rate": 9.35052631771436e-06,
+      "loss": 0.5515,
+      "step": 2969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2417818654827886,
+      "learning_rate": 9.350017791419983e-06,
+      "loss": 0.6563,
+      "step": 2970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.977732183414922,
+      "learning_rate": 9.349509079959252e-06,
+      "loss": 0.5975,
+      "step": 2971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7946625188117596,
+      "learning_rate": 9.349000183353817e-06,
+      "loss": 0.6598,
+      "step": 2972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3933140392024956,
+      "learning_rate": 9.348491101625342e-06,
+      "loss": 0.6087,
+      "step": 2973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.63265142918763,
+      "learning_rate": 9.347981834795497e-06,
+      "loss": 0.5641,
+      "step": 2974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8515643184144605,
+      "learning_rate": 9.347472382885958e-06,
+      "loss": 0.5925,
+      "step": 2975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8477336400377684,
+      "learning_rate": 9.346962745918413e-06,
+      "loss": 0.6146,
+      "step": 2976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.603407537780216,
+      "learning_rate": 9.346452923914555e-06,
+      "loss": 0.4649,
+      "step": 2977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6604501094374994,
+      "learning_rate": 9.345942916896087e-06,
+      "loss": 0.5437,
+      "step": 2978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.694988307245293,
+      "learning_rate": 9.345432724884714e-06,
+      "loss": 0.585,
+      "step": 2979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7034293684780129,
+      "learning_rate": 9.344922347902157e-06,
+      "loss": 0.5716,
+      "step": 2980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7991747189698661,
+      "learning_rate": 9.34441178597014e-06,
+      "loss": 0.5782,
+      "step": 2981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4727008222887323,
+      "learning_rate": 9.343901039110396e-06,
+      "loss": 0.6948,
+      "step": 2982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9349853240298963,
+      "learning_rate": 9.343390107344665e-06,
+      "loss": 0.6253,
+      "step": 2983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6167534352333365,
+      "learning_rate": 9.342878990694698e-06,
+      "loss": 0.587,
+      "step": 2984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7958548637169885,
+      "learning_rate": 9.34236768918225e-06,
+      "loss": 0.637,
+      "step": 2985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7553547332723676,
+      "learning_rate": 9.341856202829086e-06,
+      "loss": 0.6235,
+      "step": 2986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6116580339787628,
+      "learning_rate": 9.341344531656977e-06,
+      "loss": 0.5701,
+      "step": 2987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8274743915536293,
+      "learning_rate": 9.340832675687706e-06,
+      "loss": 0.6359,
+      "step": 2988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5895476890248806,
+      "learning_rate": 9.340320634943057e-06,
+      "loss": 0.4841,
+      "step": 2989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7026001758067324,
+      "learning_rate": 9.339808409444829e-06,
+      "loss": 0.5757,
+      "step": 2990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6487806828137881,
+      "learning_rate": 9.339295999214825e-06,
+      "loss": 0.521,
+      "step": 2991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5982423403557943,
+      "learning_rate": 9.338783404274858e-06,
+      "loss": 0.5509,
+      "step": 2992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8143840407930086,
+      "learning_rate": 9.338270624646745e-06,
+      "loss": 0.6016,
+      "step": 2993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8278309078279198,
+      "learning_rate": 9.337757660352315e-06,
+      "loss": 0.6428,
+      "step": 2994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6076523701640529,
+      "learning_rate": 9.337244511413402e-06,
+      "loss": 0.5534,
+      "step": 2995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9330251158033744,
+      "learning_rate": 9.336731177851852e-06,
+      "loss": 0.6081,
+      "step": 2996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8346508394264712,
+      "learning_rate": 9.336217659689512e-06,
+      "loss": 0.6167,
+      "step": 2997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.742525535161732,
+      "learning_rate": 9.335703956948243e-06,
+      "loss": 0.5543,
+      "step": 2998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7866807279632302,
+      "learning_rate": 9.335190069649913e-06,
+      "loss": 0.5509,
+      "step": 2999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6825555935104467,
+      "learning_rate": 9.334675997816393e-06,
+      "loss": 0.5665,
+      "step": 3000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7496473110772021,
+      "learning_rate": 9.33416174146957e-06,
+      "loss": 0.6646,
+      "step": 3001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6414747458468472,
+      "learning_rate": 9.33364730063133e-06,
+      "loss": 0.5835,
+      "step": 3002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7642314919294045,
+      "learning_rate": 9.333132675323573e-06,
+      "loss": 0.6587,
+      "step": 3003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6151487272763504,
+      "learning_rate": 9.332617865568204e-06,
+      "loss": 0.5578,
+      "step": 3004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8710684662555659,
+      "learning_rate": 9.332102871387139e-06,
+      "loss": 0.5876,
+      "step": 3005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6427377041504402,
+      "learning_rate": 9.331587692802298e-06,
+      "loss": 0.5182,
+      "step": 3006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7158472416463113,
+      "learning_rate": 9.331072329835608e-06,
+      "loss": 0.5122,
+      "step": 3007
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7418179663545695,
+      "learning_rate": 9.330556782509013e-06,
+      "loss": 0.6371,
+      "step": 3008
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5572898008872351,
+      "learning_rate": 9.330041050844451e-06,
+      "loss": 0.5122,
+      "step": 3009
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7713518982989871,
+      "learning_rate": 9.329525134863879e-06,
+      "loss": 0.6413,
+      "step": 3010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6819405797409833,
+      "learning_rate": 9.329009034589259e-06,
+      "loss": 0.553,
+      "step": 3011
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9541471572663822,
+      "learning_rate": 9.328492750042557e-06,
+      "loss": 0.6898,
+      "step": 3012
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7001677216180435,
+      "learning_rate": 9.327976281245749e-06,
+      "loss": 0.5193,
+      "step": 3013
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.696208988090597,
+      "learning_rate": 9.327459628220823e-06,
+      "loss": 0.547,
+      "step": 3014
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8183328967029198,
+      "learning_rate": 9.326942790989768e-06,
+      "loss": 0.6224,
+      "step": 3015
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8003448604103915,
+      "learning_rate": 9.326425769574586e-06,
+      "loss": 0.5968,
+      "step": 3016
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7645706263357354,
+      "learning_rate": 9.325908563997284e-06,
+      "loss": 0.5375,
+      "step": 3017
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6388165494741231,
+      "learning_rate": 9.32539117427988e-06,
+      "loss": 0.5432,
+      "step": 3018
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6579376954107429,
+      "learning_rate": 9.324873600444393e-06,
+      "loss": 0.5294,
+      "step": 3019
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5645629811200503,
+      "learning_rate": 9.324355842512858e-06,
+      "loss": 0.5076,
+      "step": 3020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7229755582677653,
+      "learning_rate": 9.323837900507313e-06,
+      "loss": 0.5746,
+      "step": 3021
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.790481285180887,
+      "learning_rate": 9.323319774449806e-06,
+      "loss": 0.6435,
+      "step": 3022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.691859607482006,
+      "learning_rate": 9.32280146436239e-06,
+      "loss": 0.6124,
+      "step": 3023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.626196159201232,
+      "learning_rate": 9.322282970267132e-06,
+      "loss": 0.5344,
+      "step": 3024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8321946396502785,
+      "learning_rate": 9.321764292186098e-06,
+      "loss": 0.6125,
+      "step": 3025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6642325621002598,
+      "learning_rate": 9.32124543014137e-06,
+      "loss": 0.5373,
+      "step": 3026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2194842676981374,
+      "learning_rate": 9.320726384155032e-06,
+      "loss": 0.6215,
+      "step": 3027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6319056284270012,
+      "learning_rate": 9.320207154249179e-06,
+      "loss": 0.5524,
+      "step": 3028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6057130280445902,
+      "learning_rate": 9.319687740445914e-06,
+      "loss": 0.5426,
+      "step": 3029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8053863670189406,
+      "learning_rate": 9.319168142767344e-06,
+      "loss": 0.6014,
+      "step": 3030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8536366098854649,
+      "learning_rate": 9.318648361235588e-06,
+      "loss": 0.5412,
+      "step": 3031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1209552390347748,
+      "learning_rate": 9.318128395872772e-06,
+      "loss": 0.6431,
+      "step": 3032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5805310978027239,
+      "learning_rate": 9.31760824670103e-06,
+      "loss": 0.5322,
+      "step": 3033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7922086275610772,
+      "learning_rate": 9.317087913742501e-06,
+      "loss": 0.5947,
+      "step": 3034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7086138941111222,
+      "learning_rate": 9.316567397019336e-06,
+      "loss": 0.4977,
+      "step": 3035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8280261185336169,
+      "learning_rate": 9.31604669655369e-06,
+      "loss": 0.639,
+      "step": 3036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.901513560973976,
+      "learning_rate": 9.315525812367728e-06,
+      "loss": 0.6625,
+      "step": 3037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6683418740981407,
+      "learning_rate": 9.315004744483623e-06,
+      "loss": 0.5734,
+      "step": 3038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5955479809766612,
+      "learning_rate": 9.314483492923555e-06,
+      "loss": 0.569,
+      "step": 3039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7098366300768894,
+      "learning_rate": 9.313962057709712e-06,
+      "loss": 0.5087,
+      "step": 3040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6644309826988876,
+      "learning_rate": 9.313440438864288e-06,
+      "loss": 0.5233,
+      "step": 3041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9052441480315545,
+      "learning_rate": 9.312918636409488e-06,
+      "loss": 0.5835,
+      "step": 3042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7402182307744731,
+      "learning_rate": 9.312396650367528e-06,
+      "loss": 0.6059,
+      "step": 3043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5881936680543857,
+      "learning_rate": 9.311874480760619e-06,
+      "loss": 0.4582,
+      "step": 3044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7831489203537119,
+      "learning_rate": 9.311352127610995e-06,
+      "loss": 0.648,
+      "step": 3045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5943426347788903,
+      "learning_rate": 9.310829590940886e-06,
+      "loss": 0.5038,
+      "step": 3046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6626827551503282,
+      "learning_rate": 9.310306870772536e-06,
+      "loss": 0.6143,
+      "step": 3047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6808381714363823,
+      "learning_rate": 9.3097839671282e-06,
+      "loss": 0.5389,
+      "step": 3048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7441243592991102,
+      "learning_rate": 9.309260880030128e-06,
+      "loss": 0.5767,
+      "step": 3049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6852853431179967,
+      "learning_rate": 9.308737609500593e-06,
+      "loss": 0.5732,
+      "step": 3050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8530760941007216,
+      "learning_rate": 9.308214155561866e-06,
+      "loss": 0.6467,
+      "step": 3051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6322104196931244,
+      "learning_rate": 9.30769051823623e-06,
+      "loss": 0.4893,
+      "step": 3052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1032182049391803,
+      "learning_rate": 9.307166697545976e-06,
+      "loss": 0.6418,
+      "step": 3053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.546218845958658,
+      "learning_rate": 9.306642693513397e-06,
+      "loss": 0.5108,
+      "step": 3054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5662008064598786,
+      "learning_rate": 9.3061185061608e-06,
+      "loss": 0.5168,
+      "step": 3055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7800336009647092,
+      "learning_rate": 9.3055941355105e-06,
+      "loss": 0.6175,
+      "step": 3056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7333866231083132,
+      "learning_rate": 9.305069581584816e-06,
+      "loss": 0.5589,
+      "step": 3057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6769398353688174,
+      "learning_rate": 9.304544844406077e-06,
+      "loss": 0.5824,
+      "step": 3058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6581380436974328,
+      "learning_rate": 9.304019923996619e-06,
+      "loss": 0.5922,
+      "step": 3059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6768552942287963,
+      "learning_rate": 9.303494820378787e-06,
+      "loss": 0.6227,
+      "step": 3060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6824732779233269,
+      "learning_rate": 9.302969533574933e-06,
+      "loss": 0.5432,
+      "step": 3061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9017951473668746,
+      "learning_rate": 9.302444063607415e-06,
+      "loss": 0.6175,
+      "step": 3062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9157586394484665,
+      "learning_rate": 9.301918410498603e-06,
+      "loss": 0.6712,
+      "step": 3063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7766878050601725,
+      "learning_rate": 9.301392574270871e-06,
+      "loss": 0.6945,
+      "step": 3064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6717405829253689,
+      "learning_rate": 9.300866554946601e-06,
+      "loss": 0.5873,
+      "step": 3065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7471310858999942,
+      "learning_rate": 9.300340352548187e-06,
+      "loss": 0.5998,
+      "step": 3066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7138892208972915,
+      "learning_rate": 9.299813967098025e-06,
+      "loss": 0.5422,
+      "step": 3067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7148621043066545,
+      "learning_rate": 9.299287398618523e-06,
+      "loss": 0.5343,
+      "step": 3068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5276683100930166,
+      "learning_rate": 9.298760647132096e-06,
+      "loss": 0.4901,
+      "step": 3069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6157281448589456,
+      "learning_rate": 9.298233712661166e-06,
+      "loss": 0.5509,
+      "step": 3070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5890621180780182,
+      "learning_rate": 9.29770659522816e-06,
+      "loss": 0.536,
+      "step": 3071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5949375071132643,
+      "learning_rate": 9.297179294855519e-06,
+      "loss": 0.5854,
+      "step": 3072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7321215936111286,
+      "learning_rate": 9.296651811565685e-06,
+      "loss": 0.6554,
+      "step": 3073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6632960728330101,
+      "learning_rate": 9.296124145381116e-06,
+      "loss": 0.5441,
+      "step": 3074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8927654358690954,
+      "learning_rate": 9.295596296324268e-06,
+      "loss": 0.5433,
+      "step": 3075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5917597025145906,
+      "learning_rate": 9.295068264417615e-06,
+      "loss": 0.5124,
+      "step": 3076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8730540277709486,
+      "learning_rate": 9.294540049683629e-06,
+      "loss": 0.6292,
+      "step": 3077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.76195088902274,
+      "learning_rate": 9.2940116521448e-06,
+      "loss": 0.6291,
+      "step": 3078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7144567401257371,
+      "learning_rate": 9.293483071823612e-06,
+      "loss": 0.5584,
+      "step": 3079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8238836202252123,
+      "learning_rate": 9.292954308742572e-06,
+      "loss": 0.6001,
+      "step": 3080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5557593549520737,
+      "learning_rate": 9.292425362924185e-06,
+      "loss": 0.5021,
+      "step": 3081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6445106737516632,
+      "learning_rate": 9.291896234390966e-06,
+      "loss": 0.5657,
+      "step": 3082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7875561067619063,
+      "learning_rate": 9.291366923165442e-06,
+      "loss": 0.562,
+      "step": 3083
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5296996203046099,
+      "learning_rate": 9.290837429270138e-06,
+      "loss": 0.4997,
+      "step": 3084
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7394728089933901,
+      "learning_rate": 9.290307752727598e-06,
+      "loss": 0.5575,
+      "step": 3085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.712067102744521,
+      "learning_rate": 9.289777893560368e-06,
+      "loss": 0.5479,
+      "step": 3086
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7463792735340825,
+      "learning_rate": 9.289247851791e-06,
+      "loss": 0.5827,
+      "step": 3087
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6464882057670299,
+      "learning_rate": 9.288717627442056e-06,
+      "loss": 0.571,
+      "step": 3088
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8399134938837616,
+      "learning_rate": 9.288187220536108e-06,
+      "loss": 0.6561,
+      "step": 3089
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6463657899447613,
+      "learning_rate": 9.287656631095737e-06,
+      "loss": 0.554,
+      "step": 3090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6640107311547536,
+      "learning_rate": 9.287125859143519e-06,
+      "loss": 0.5583,
+      "step": 3091
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6834728724945228,
+      "learning_rate": 9.286594904702056e-06,
+      "loss": 0.5647,
+      "step": 3092
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7438160033033764,
+      "learning_rate": 9.286063767793946e-06,
+      "loss": 0.5557,
+      "step": 3093
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7879538753250512,
+      "learning_rate": 9.285532448441796e-06,
+      "loss": 0.6375,
+      "step": 3094
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5826151051188484,
+      "learning_rate": 9.285000946668225e-06,
+      "loss": 0.5366,
+      "step": 3095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8468259040440832,
+      "learning_rate": 9.284469262495858e-06,
+      "loss": 0.6204,
+      "step": 3096
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8301424864486188,
+      "learning_rate": 9.283937395947324e-06,
+      "loss": 0.545,
+      "step": 3097
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5769045292234983,
+      "learning_rate": 9.283405347045265e-06,
+      "loss": 0.5145,
+      "step": 3098
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7910485700784748,
+      "learning_rate": 9.282873115812328e-06,
+      "loss": 0.6243,
+      "step": 3099
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7994343145544429,
+      "learning_rate": 9.28234070227117e-06,
+      "loss": 0.6042,
+      "step": 3100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7675790930144175,
+      "learning_rate": 9.281808106444452e-06,
+      "loss": 0.5719,
+      "step": 3101
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7582156493616083,
+      "learning_rate": 9.281275328354845e-06,
+      "loss": 0.6003,
+      "step": 3102
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5871920321578749,
+      "learning_rate": 9.280742368025027e-06,
+      "loss": 0.575,
+      "step": 3103
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5610931006296737,
+      "learning_rate": 9.280209225477688e-06,
+      "loss": 0.5215,
+      "step": 3104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7206658624226153,
+      "learning_rate": 9.279675900735519e-06,
+      "loss": 0.6207,
+      "step": 3105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6770833910428605,
+      "learning_rate": 9.279142393821224e-06,
+      "loss": 0.5367,
+      "step": 3106
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8784298211966007,
+      "learning_rate": 9.278608704757509e-06,
+      "loss": 0.6006,
+      "step": 3107
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0511083295817143,
+      "learning_rate": 9.278074833567094e-06,
+      "loss": 0.635,
+      "step": 3108
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7725134460781607,
+      "learning_rate": 9.277540780272707e-06,
+      "loss": 0.66,
+      "step": 3109
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9074875949708152,
+      "learning_rate": 9.277006544897075e-06,
+      "loss": 0.6412,
+      "step": 3110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5747213986561281,
+      "learning_rate": 9.276472127462943e-06,
+      "loss": 0.5407,
+      "step": 3111
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6084753742855469,
+      "learning_rate": 9.275937527993058e-06,
+      "loss": 0.5673,
+      "step": 3112
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7364418087309035,
+      "learning_rate": 9.275402746510179e-06,
+      "loss": 0.5565,
+      "step": 3113
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6074742736840726,
+      "learning_rate": 9.274867783037063e-06,
+      "loss": 0.4966,
+      "step": 3114
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7712402099091962,
+      "learning_rate": 9.274332637596488e-06,
+      "loss": 0.6442,
+      "step": 3115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5292188506702172,
+      "learning_rate": 9.273797310211232e-06,
+      "loss": 0.4742,
+      "step": 3116
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7506883236759608,
+      "learning_rate": 9.273261800904082e-06,
+      "loss": 0.5994,
+      "step": 3117
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8367115217354989,
+      "learning_rate": 9.272726109697829e-06,
+      "loss": 0.6219,
+      "step": 3118
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6130562197922128,
+      "learning_rate": 9.272190236615283e-06,
+      "loss": 0.4886,
+      "step": 3119
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6482109192490533,
+      "learning_rate": 9.271654181679249e-06,
+      "loss": 0.5192,
+      "step": 3120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6525488995115156,
+      "learning_rate": 9.271117944912546e-06,
+      "loss": 0.5351,
+      "step": 3121
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8013882323655882,
+      "learning_rate": 9.270581526338003e-06,
+      "loss": 0.5881,
+      "step": 3122
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8160372182766578,
+      "learning_rate": 9.27004492597845e-06,
+      "loss": 0.6463,
+      "step": 3123
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7607549688946318,
+      "learning_rate": 9.269508143856728e-06,
+      "loss": 0.6309,
+      "step": 3124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7238568313130813,
+      "learning_rate": 9.26897117999569e-06,
+      "loss": 0.6006,
+      "step": 3125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6115375378503345,
+      "learning_rate": 9.26843403441819e-06,
+      "loss": 0.5531,
+      "step": 3126
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9119037048165477,
+      "learning_rate": 9.267896707147093e-06,
+      "loss": 0.6225,
+      "step": 3127
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6926218128611051,
+      "learning_rate": 9.26735919820527e-06,
+      "loss": 0.5986,
+      "step": 3128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8324995338310999,
+      "learning_rate": 9.266821507615605e-06,
+      "loss": 0.6328,
+      "step": 3129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8445941146845337,
+      "learning_rate": 9.266283635400981e-06,
+      "loss": 0.6779,
+      "step": 3130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6201031681120114,
+      "learning_rate": 9.265745581584297e-06,
+      "loss": 0.5428,
+      "step": 3131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8656280762912113,
+      "learning_rate": 9.265207346188456e-06,
+      "loss": 0.6874,
+      "step": 3132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8283757591201077,
+      "learning_rate": 9.264668929236365e-06,
+      "loss": 0.5856,
+      "step": 3133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5957631667069805,
+      "learning_rate": 9.264130330750949e-06,
+      "loss": 0.5381,
+      "step": 3134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7690889525596007,
+      "learning_rate": 9.26359155075513e-06,
+      "loss": 0.6318,
+      "step": 3135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7993677063373822,
+      "learning_rate": 9.263052589271841e-06,
+      "loss": 0.6209,
+      "step": 3136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6530563933064154,
+      "learning_rate": 9.262513446324027e-06,
+      "loss": 0.5409,
+      "step": 3137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9396208535520592,
+      "learning_rate": 9.261974121934637e-06,
+      "loss": 0.6181,
+      "step": 3138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5817774325248966,
+      "learning_rate": 9.261434616126628e-06,
+      "loss": 0.5174,
+      "step": 3139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8179224407841669,
+      "learning_rate": 9.260894928922966e-06,
+      "loss": 0.5769,
+      "step": 3140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6675953013812651,
+      "learning_rate": 9.260355060346622e-06,
+      "loss": 0.5401,
+      "step": 3141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6047063955764763,
+      "learning_rate": 9.259815010420577e-06,
+      "loss": 0.5212,
+      "step": 3142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8137771625454581,
+      "learning_rate": 9.259274779167819e-06,
+      "loss": 0.5973,
+      "step": 3143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5891073724621664,
+      "learning_rate": 9.258734366611345e-06,
+      "loss": 0.4764,
+      "step": 3144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6936274196117064,
+      "learning_rate": 9.258193772774158e-06,
+      "loss": 0.5844,
+      "step": 3145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6005108952985638,
+      "learning_rate": 9.257652997679268e-06,
+      "loss": 0.5612,
+      "step": 3146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.677342080236454,
+      "learning_rate": 9.257112041349697e-06,
+      "loss": 0.547,
+      "step": 3147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6082332538703317,
+      "learning_rate": 9.25657090380847e-06,
+      "loss": 0.5582,
+      "step": 3148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8679094395986775,
+      "learning_rate": 9.256029585078622e-06,
+      "loss": 0.6566,
+      "step": 3149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6503681281399057,
+      "learning_rate": 9.255488085183193e-06,
+      "loss": 0.5036,
+      "step": 3150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7907090344306829,
+      "learning_rate": 9.254946404145236e-06,
+      "loss": 0.6045,
+      "step": 3151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8024687216639622,
+      "learning_rate": 9.254404541987807e-06,
+      "loss": 0.645,
+      "step": 3152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5642301377950192,
+      "learning_rate": 9.253862498733972e-06,
+      "loss": 0.535,
+      "step": 3153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7766938738624718,
+      "learning_rate": 9.253320274406806e-06,
+      "loss": 0.5819,
+      "step": 3154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6637008314505372,
+      "learning_rate": 9.252777869029386e-06,
+      "loss": 0.5625,
+      "step": 3155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7747869499782861,
+      "learning_rate": 9.252235282624801e-06,
+      "loss": 0.594,
+      "step": 3156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6668475851783979,
+      "learning_rate": 9.251692515216148e-06,
+      "loss": 0.5273,
+      "step": 3157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8544566294686278,
+      "learning_rate": 9.251149566826533e-06,
+      "loss": 0.5744,
+      "step": 3158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.781423719362171,
+      "learning_rate": 9.250606437479064e-06,
+      "loss": 0.5431,
+      "step": 3159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6466778267197053,
+      "learning_rate": 9.250063127196863e-06,
+      "loss": 0.5671,
+      "step": 3160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7482778024640498,
+      "learning_rate": 9.249519636003056e-06,
+      "loss": 0.5943,
+      "step": 3161
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6519006468416555,
+      "learning_rate": 9.248975963920777e-06,
+      "loss": 0.549,
+      "step": 3162
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6622251030234204,
+      "learning_rate": 9.24843211097317e-06,
+      "loss": 0.5245,
+      "step": 3163
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7375639692494137,
+      "learning_rate": 9.247888077183382e-06,
+      "loss": 0.5534,
+      "step": 3164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7985924562798349,
+      "learning_rate": 9.247343862574575e-06,
+      "loss": 0.6107,
+      "step": 3165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5996574487719852,
+      "learning_rate": 9.24679946716991e-06,
+      "loss": 0.4911,
+      "step": 3166
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5761423746533978,
+      "learning_rate": 9.246254890992565e-06,
+      "loss": 0.5138,
+      "step": 3167
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7984236349512767,
+      "learning_rate": 9.245710134065718e-06,
+      "loss": 0.6416,
+      "step": 3168
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7003687244052177,
+      "learning_rate": 9.245165196412558e-06,
+      "loss": 0.5389,
+      "step": 3169
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.158307663065942,
+      "learning_rate": 9.244620078056281e-06,
+      "loss": 0.5646,
+      "step": 3170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7284331813544216,
+      "learning_rate": 9.244074779020088e-06,
+      "loss": 0.6357,
+      "step": 3171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6903426199710162,
+      "learning_rate": 9.243529299327198e-06,
+      "loss": 0.5638,
+      "step": 3172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6913996326403633,
+      "learning_rate": 9.242983639000825e-06,
+      "loss": 0.6357,
+      "step": 3173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8558684031398051,
+      "learning_rate": 9.242437798064197e-06,
+      "loss": 0.5699,
+      "step": 3174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6233453702160608,
+      "learning_rate": 9.241891776540549e-06,
+      "loss": 0.562,
+      "step": 3175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7623320687446391,
+      "learning_rate": 9.241345574453122e-06,
+      "loss": 0.6822,
+      "step": 3176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8212404104712616,
+      "learning_rate": 9.240799191825167e-06,
+      "loss": 0.6634,
+      "step": 3177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6892695539055483,
+      "learning_rate": 9.240252628679945e-06,
+      "loss": 0.5825,
+      "step": 3178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7474256591434516,
+      "learning_rate": 9.239705885040716e-06,
+      "loss": 0.5885,
+      "step": 3179
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.701528282681911,
+      "learning_rate": 9.239158960930757e-06,
+      "loss": 0.5961,
+      "step": 3180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.614972457800695,
+      "learning_rate": 9.238611856373346e-06,
+      "loss": 0.5648,
+      "step": 3181
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7016096696878144,
+      "learning_rate": 9.238064571391772e-06,
+      "loss": 0.5871,
+      "step": 3182
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6065021393621388,
+      "learning_rate": 9.237517106009335e-06,
+      "loss": 0.5456,
+      "step": 3183
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6165816396422367,
+      "learning_rate": 9.236969460249335e-06,
+      "loss": 0.4853,
+      "step": 3184
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9406136898950629,
+      "learning_rate": 9.236421634135085e-06,
+      "loss": 0.6237,
+      "step": 3185
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7294982034530778,
+      "learning_rate": 9.235873627689902e-06,
+      "loss": 0.5752,
+      "step": 3186
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6475248770643098,
+      "learning_rate": 9.235325440937115e-06,
+      "loss": 0.5316,
+      "step": 3187
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8161288507107136,
+      "learning_rate": 9.234777073900059e-06,
+      "loss": 0.6273,
+      "step": 3188
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9524174405826741,
+      "learning_rate": 9.234228526602074e-06,
+      "loss": 0.6585,
+      "step": 3189
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7622085834034301,
+      "learning_rate": 9.233679799066514e-06,
+      "loss": 0.5887,
+      "step": 3190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8001198885334686,
+      "learning_rate": 9.233130891316731e-06,
+      "loss": 0.6415,
+      "step": 3191
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7299578815119844,
+      "learning_rate": 9.232581803376095e-06,
+      "loss": 0.6307,
+      "step": 3192
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7906144659315226,
+      "learning_rate": 9.232032535267976e-06,
+      "loss": 0.6024,
+      "step": 3193
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7027847323877751,
+      "learning_rate": 9.231483087015755e-06,
+      "loss": 0.6417,
+      "step": 3194
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8302817529568468,
+      "learning_rate": 9.230933458642823e-06,
+      "loss": 0.562,
+      "step": 3195
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8739537638604074,
+      "learning_rate": 9.230383650172572e-06,
+      "loss": 0.6126,
+      "step": 3196
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5171971506646962,
+      "learning_rate": 9.229833661628409e-06,
+      "loss": 0.4514,
+      "step": 3197
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7834864512856085,
+      "learning_rate": 9.229283493033742e-06,
+      "loss": 0.621,
+      "step": 3198
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6275152321814564,
+      "learning_rate": 9.228733144411993e-06,
+      "loss": 0.5007,
+      "step": 3199
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1290054452156002,
+      "learning_rate": 9.228182615786585e-06,
+      "loss": 0.6702,
+      "step": 3200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8395437716995005,
+      "learning_rate": 9.227631907180956e-06,
+      "loss": 0.5648,
+      "step": 3201
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7740440337826754,
+      "learning_rate": 9.227081018618546e-06,
+      "loss": 0.6478,
+      "step": 3202
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7305535433094422,
+      "learning_rate": 9.226529950122805e-06,
+      "loss": 0.5504,
+      "step": 3203
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5646711821983378,
+      "learning_rate": 9.225978701717189e-06,
+      "loss": 0.5423,
+      "step": 3204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6183282577703919,
+      "learning_rate": 9.225427273425165e-06,
+      "loss": 0.5615,
+      "step": 3205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6229665852162618,
+      "learning_rate": 9.224875665270206e-06,
+      "loss": 0.5171,
+      "step": 3206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5431499851299624,
+      "learning_rate": 9.224323877275786e-06,
+      "loss": 0.5199,
+      "step": 3207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6601271472318798,
+      "learning_rate": 9.223771909465402e-06,
+      "loss": 0.5443,
+      "step": 3208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6404625223145297,
+      "learning_rate": 9.223219761862545e-06,
+      "loss": 0.5509,
+      "step": 3209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7037978654532602,
+      "learning_rate": 9.222667434490716e-06,
+      "loss": 0.6473,
+      "step": 3210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.721589651611027,
+      "learning_rate": 9.222114927373429e-06,
+      "loss": 0.5679,
+      "step": 3211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6568083250973575,
+      "learning_rate": 9.2215622405342e-06,
+      "loss": 0.5928,
+      "step": 3212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7164926491186786,
+      "learning_rate": 9.22100937399656e-06,
+      "loss": 0.6043,
+      "step": 3213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8310340783994664,
+      "learning_rate": 9.220456327784038e-06,
+      "loss": 0.6041,
+      "step": 3214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7426404072208325,
+      "learning_rate": 9.219903101920176e-06,
+      "loss": 0.6161,
+      "step": 3215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8143939296303716,
+      "learning_rate": 9.219349696428523e-06,
+      "loss": 0.6158,
+      "step": 3216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7328724560042611,
+      "learning_rate": 9.218796111332638e-06,
+      "loss": 0.5665,
+      "step": 3217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.014845447845572,
+      "learning_rate": 9.218242346656083e-06,
+      "loss": 0.625,
+      "step": 3218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9705993139030301,
+      "learning_rate": 9.217688402422431e-06,
+      "loss": 0.6891,
+      "step": 3219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7499917739142293,
+      "learning_rate": 9.217134278655262e-06,
+      "loss": 0.5755,
+      "step": 3220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7125004999350245,
+      "learning_rate": 9.216579975378163e-06,
+      "loss": 0.5191,
+      "step": 3221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.854810055587463,
+      "learning_rate": 9.21602549261473e-06,
+      "loss": 0.6089,
+      "step": 3222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5691407539342304,
+      "learning_rate": 9.215470830388562e-06,
+      "loss": 0.5109,
+      "step": 3223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6172279238982875,
+      "learning_rate": 9.21491598872327e-06,
+      "loss": 0.5294,
+      "step": 3224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7981602876094932,
+      "learning_rate": 9.214360967642477e-06,
+      "loss": 0.6234,
+      "step": 3225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6538744727864495,
+      "learning_rate": 9.213805767169802e-06,
+      "loss": 0.5342,
+      "step": 3226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7244812345654913,
+      "learning_rate": 9.213250387328882e-06,
+      "loss": 0.5867,
+      "step": 3227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7182566480602726,
+      "learning_rate": 9.212694828143356e-06,
+      "loss": 0.5795,
+      "step": 3228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8816372560725132,
+      "learning_rate": 9.212139089636875e-06,
+      "loss": 0.627,
+      "step": 3229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5951494522354634,
+      "learning_rate": 9.211583171833092e-06,
+      "loss": 0.5791,
+      "step": 3230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6931626072454565,
+      "learning_rate": 9.211027074755671e-06,
+      "loss": 0.5897,
+      "step": 3231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6833677451982374,
+      "learning_rate": 9.210470798428283e-06,
+      "loss": 0.5215,
+      "step": 3232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.282989705730721,
+      "learning_rate": 9.20991434287461e-06,
+      "loss": 0.6345,
+      "step": 3233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5855130011417814,
+      "learning_rate": 9.209357708118335e-06,
+      "loss": 0.5138,
+      "step": 3234
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.596196516944751,
+      "learning_rate": 9.208800894183154e-06,
+      "loss": 0.5724,
+      "step": 3235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7480587414485678,
+      "learning_rate": 9.208243901092769e-06,
+      "loss": 0.6149,
+      "step": 3236
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8518895744305718,
+      "learning_rate": 9.207686728870889e-06,
+      "loss": 0.683,
+      "step": 3237
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.892959194286907,
+      "learning_rate": 9.207129377541229e-06,
+      "loss": 0.6175,
+      "step": 3238
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5810982847496905,
+      "learning_rate": 9.206571847127517e-06,
+      "loss": 0.5218,
+      "step": 3239
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8535470654985331,
+      "learning_rate": 9.206014137653482e-06,
+      "loss": 0.6381,
+      "step": 3240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.633664919505574,
+      "learning_rate": 9.205456249142868e-06,
+      "loss": 0.555,
+      "step": 3241
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8160371523072273,
+      "learning_rate": 9.204898181619419e-06,
+      "loss": 0.5539,
+      "step": 3242
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7079567478379215,
+      "learning_rate": 9.204339935106892e-06,
+      "loss": 0.599,
+      "step": 3243
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7615625775965033,
+      "learning_rate": 9.203781509629048e-06,
+      "loss": 0.5952,
+      "step": 3244
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7544846974124569,
+      "learning_rate": 9.203222905209659e-06,
+      "loss": 0.6333,
+      "step": 3245
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6298341210044557,
+      "learning_rate": 9.202664121872502e-06,
+      "loss": 0.5374,
+      "step": 3246
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5971652832986465,
+      "learning_rate": 9.202105159641363e-06,
+      "loss": 0.5817,
+      "step": 3247
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7216286265188634,
+      "learning_rate": 9.201546018540037e-06,
+      "loss": 0.509,
+      "step": 3248
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8146484420358,
+      "learning_rate": 9.200986698592323e-06,
+      "loss": 0.6646,
+      "step": 3249
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6470741855321337,
+      "learning_rate": 9.200427199822028e-06,
+      "loss": 0.5547,
+      "step": 3250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5307516565538946,
+      "learning_rate": 9.19986752225297e-06,
+      "loss": 0.4952,
+      "step": 3251
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7499313470267501,
+      "learning_rate": 9.199307665908974e-06,
+      "loss": 0.6223,
+      "step": 3252
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7873693261187309,
+      "learning_rate": 9.19874763081387e-06,
+      "loss": 0.5919,
+      "step": 3253
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8312459521266437,
+      "learning_rate": 9.198187416991493e-06,
+      "loss": 0.5581,
+      "step": 3254
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6218252502503935,
+      "learning_rate": 9.197627024465697e-06,
+      "loss": 0.5271,
+      "step": 3255
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6793659870096235,
+      "learning_rate": 9.197066453260331e-06,
+      "loss": 0.5878,
+      "step": 3256
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8097458712161544,
+      "learning_rate": 9.196505703399257e-06,
+      "loss": 0.6745,
+      "step": 3257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8928733149820136,
+      "learning_rate": 9.195944774906347e-06,
+      "loss": 0.6213,
+      "step": 3258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6472409547449489,
+      "learning_rate": 9.195383667805475e-06,
+      "loss": 0.5767,
+      "step": 3259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6662930157629626,
+      "learning_rate": 9.194822382120527e-06,
+      "loss": 0.5626,
+      "step": 3260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9905387848773961,
+      "learning_rate": 9.194260917875395e-06,
+      "loss": 0.635,
+      "step": 3261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6018491781279309,
+      "learning_rate": 9.193699275093979e-06,
+      "loss": 0.5283,
+      "step": 3262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7426946739996282,
+      "learning_rate": 9.193137453800185e-06,
+      "loss": 0.5722,
+      "step": 3263
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5828078313417627,
+      "learning_rate": 9.19257545401793e-06,
+      "loss": 0.5268,
+      "step": 3264
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9124170137727117,
+      "learning_rate": 9.192013275771136e-06,
+      "loss": 0.6762,
+      "step": 3265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7039819945593421,
+      "learning_rate": 9.191450919083731e-06,
+      "loss": 0.5918,
+      "step": 3266
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7032769511281745,
+      "learning_rate": 9.190888383979655e-06,
+      "loss": 0.5774,
+      "step": 3267
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7520639233199657,
+      "learning_rate": 9.190325670482854e-06,
+      "loss": 0.6654,
+      "step": 3268
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6459007792468826,
+      "learning_rate": 9.189762778617277e-06,
+      "loss": 0.552,
+      "step": 3269
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7001160285666483,
+      "learning_rate": 9.189199708406889e-06,
+      "loss": 0.5639,
+      "step": 3270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6140801763175019,
+      "learning_rate": 9.188636459875655e-06,
+      "loss": 0.5589,
+      "step": 3271
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.153917460766353,
+      "learning_rate": 9.188073033047553e-06,
+      "loss": 0.6643,
+      "step": 3272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7246733619528902,
+      "learning_rate": 9.187509427946565e-06,
+      "loss": 0.5491,
+      "step": 3273
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5658520188236634,
+      "learning_rate": 9.18694564459668e-06,
+      "loss": 0.5345,
+      "step": 3274
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8130835488751036,
+      "learning_rate": 9.1863816830219e-06,
+      "loss": 0.5984,
+      "step": 3275
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5737981169444385,
+      "learning_rate": 9.185817543246232e-06,
+      "loss": 0.5141,
+      "step": 3276
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6484394591916782,
+      "learning_rate": 9.185253225293684e-06,
+      "loss": 0.574,
+      "step": 3277
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8387932616442788,
+      "learning_rate": 9.184688729188281e-06,
+      "loss": 0.5799,
+      "step": 3278
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6059481311429133,
+      "learning_rate": 9.18412405495405e-06,
+      "loss": 0.4868,
+      "step": 3279
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5912994012192516,
+      "learning_rate": 9.18355920261503e-06,
+      "loss": 0.5601,
+      "step": 3280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6053788989091289,
+      "learning_rate": 9.182994172195266e-06,
+      "loss": 0.5473,
+      "step": 3281
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6659146771005405,
+      "learning_rate": 9.182428963718804e-06,
+      "loss": 0.5679,
+      "step": 3282
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8035257479160005,
+      "learning_rate": 9.181863577209705e-06,
+      "loss": 0.6232,
+      "step": 3283
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8270559144867194,
+      "learning_rate": 9.181298012692039e-06,
+      "loss": 0.6617,
+      "step": 3284
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.731995430929404,
+      "learning_rate": 9.180732270189877e-06,
+      "loss": 0.5651,
+      "step": 3285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6317926449243199,
+      "learning_rate": 9.180166349727303e-06,
+      "loss": 0.542,
+      "step": 3286
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.74161766044179,
+      "learning_rate": 9.179600251328405e-06,
+      "loss": 0.6183,
+      "step": 3287
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6565143061279405,
+      "learning_rate": 9.17903397501728e-06,
+      "loss": 0.5836,
+      "step": 3288
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9126999944088788,
+      "learning_rate": 9.178467520818033e-06,
+      "loss": 0.6003,
+      "step": 3289
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8470934289346624,
+      "learning_rate": 9.177900888754775e-06,
+      "loss": 0.6555,
+      "step": 3290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6351474911842814,
+      "learning_rate": 9.17733407885163e-06,
+      "loss": 0.5221,
+      "step": 3291
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6176272144988341,
+      "learning_rate": 9.176767091132718e-06,
+      "loss": 0.5505,
+      "step": 3292
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5704415893873597,
+      "learning_rate": 9.176199925622183e-06,
+      "loss": 0.5121,
+      "step": 3293
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6747566530757642,
+      "learning_rate": 9.175632582344158e-06,
+      "loss": 0.4967,
+      "step": 3294
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7974450169225238,
+      "learning_rate": 9.175065061322799e-06,
+      "loss": 0.6486,
+      "step": 3295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5971550840646895,
+      "learning_rate": 9.174497362582262e-06,
+      "loss": 0.5224,
+      "step": 3296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8160735889062235,
+      "learning_rate": 9.173929486146713e-06,
+      "loss": 0.637,
+      "step": 3297
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7745251199837998,
+      "learning_rate": 9.173361432040321e-06,
+      "loss": 0.6084,
+      "step": 3298
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6976358887644472,
+      "learning_rate": 9.172793200287272e-06,
+      "loss": 0.5466,
+      "step": 3299
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.64552829772451,
+      "learning_rate": 9.172224790911748e-06,
+      "loss": 0.5751,
+      "step": 3300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6585268850253904,
+      "learning_rate": 9.17165620393795e-06,
+      "loss": 0.6338,
+      "step": 3301
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6721880760623039,
+      "learning_rate": 9.171087439390076e-06,
+      "loss": 0.5549,
+      "step": 3302
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5779134645775404,
+      "learning_rate": 9.17051849729234e-06,
+      "loss": 0.4685,
+      "step": 3303
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9383967938567331,
+      "learning_rate": 9.169949377668958e-06,
+      "loss": 0.7189,
+      "step": 3304
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.618739152947225,
+      "learning_rate": 9.169380080544157e-06,
+      "loss": 0.5339,
+      "step": 3305
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6074685136042486,
+      "learning_rate": 9.168810605942168e-06,
+      "loss": 0.5385,
+      "step": 3306
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3964853215673774,
+      "learning_rate": 9.168240953887234e-06,
+      "loss": 0.6239,
+      "step": 3307
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8854781551920204,
+      "learning_rate": 9.167671124403601e-06,
+      "loss": 0.6395,
+      "step": 3308
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6341463647232668,
+      "learning_rate": 9.167101117515528e-06,
+      "loss": 0.5084,
+      "step": 3309
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6204243305855078,
+      "learning_rate": 9.166530933247276e-06,
+      "loss": 0.4837,
+      "step": 3310
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8782656556572855,
+      "learning_rate": 9.165960571623116e-06,
+      "loss": 0.6147,
+      "step": 3311
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5921902972365098,
+      "learning_rate": 9.165390032667328e-06,
+      "loss": 0.5314,
+      "step": 3312
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6168798103946963,
+      "learning_rate": 9.164819316404197e-06,
+      "loss": 0.549,
+      "step": 3313
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6546515366806996,
+      "learning_rate": 9.164248422858015e-06,
+      "loss": 0.5533,
+      "step": 3314
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.527765102681221,
+      "learning_rate": 9.163677352053087e-06,
+      "loss": 0.4784,
+      "step": 3315
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5243501757772494,
+      "learning_rate": 9.163106104013717e-06,
+      "loss": 0.5187,
+      "step": 3316
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8512706500980287,
+      "learning_rate": 9.162534678764225e-06,
+      "loss": 0.6358,
+      "step": 3317
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.801562613911495,
+      "learning_rate": 9.161963076328933e-06,
+      "loss": 0.6723,
+      "step": 3318
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9063120931367663,
+      "learning_rate": 9.161391296732171e-06,
+      "loss": 0.6395,
+      "step": 3319
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6613677803151039,
+      "learning_rate": 9.160819339998282e-06,
+      "loss": 0.5656,
+      "step": 3320
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8788521624452991,
+      "learning_rate": 9.160247206151608e-06,
+      "loss": 0.6328,
+      "step": 3321
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9102758162586393,
+      "learning_rate": 9.159674895216507e-06,
+      "loss": 0.6288,
+      "step": 3322
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6292348531350221,
+      "learning_rate": 9.159102407217334e-06,
+      "loss": 0.5138,
+      "step": 3323
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6926993439377039,
+      "learning_rate": 9.158529742178466e-06,
+      "loss": 0.5795,
+      "step": 3324
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8196574769217888,
+      "learning_rate": 9.157956900124273e-06,
+      "loss": 0.6138,
+      "step": 3325
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6584602786355435,
+      "learning_rate": 9.157383881079143e-06,
+      "loss": 0.5792,
+      "step": 3326
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0104669688959764,
+      "learning_rate": 9.156810685067467e-06,
+      "loss": 0.6747,
+      "step": 3327
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6532782793076529,
+      "learning_rate": 9.156237312113642e-06,
+      "loss": 0.5303,
+      "step": 3328
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8287066372914469,
+      "learning_rate": 9.155663762242076e-06,
+      "loss": 0.6336,
+      "step": 3329
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6742819142673648,
+      "learning_rate": 9.155090035477185e-06,
+      "loss": 0.579,
+      "step": 3330
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.953562904369008,
+      "learning_rate": 9.154516131843388e-06,
+      "loss": 0.6716,
+      "step": 3331
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6873147643866758,
+      "learning_rate": 9.153942051365115e-06,
+      "loss": 0.5199,
+      "step": 3332
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9029028758271641,
+      "learning_rate": 9.153367794066802e-06,
+      "loss": 0.5479,
+      "step": 3333
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6772567790207706,
+      "learning_rate": 9.152793359972894e-06,
+      "loss": 0.5512,
+      "step": 3334
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.846653051375728,
+      "learning_rate": 9.152218749107845e-06,
+      "loss": 0.6476,
+      "step": 3335
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.624086643969705,
+      "learning_rate": 9.151643961496111e-06,
+      "loss": 0.575,
+      "step": 3336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9148599476333431,
+      "learning_rate": 9.151068997162163e-06,
+      "loss": 0.6742,
+      "step": 3337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5664197078715828,
+      "learning_rate": 9.15049385613047e-06,
+      "loss": 0.5124,
+      "step": 3338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6044488824927788,
+      "learning_rate": 9.149918538425517e-06,
+      "loss": 0.5296,
+      "step": 3339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5886124520905012,
+      "learning_rate": 9.149343044071793e-06,
+      "loss": 0.5609,
+      "step": 3340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5699769180161598,
+      "learning_rate": 9.148767373093796e-06,
+      "loss": 0.5536,
+      "step": 3341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5924989604629504,
+      "learning_rate": 9.148191525516027e-06,
+      "loss": 0.5398,
+      "step": 3342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7687488774906683,
+      "learning_rate": 9.147615501363004e-06,
+      "loss": 0.5943,
+      "step": 3343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5523382658679526,
+      "learning_rate": 9.147039300659241e-06,
+      "loss": 0.5572,
+      "step": 3344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7726880357411495,
+      "learning_rate": 9.146462923429269e-06,
+      "loss": 0.6564,
+      "step": 3345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5975929958069081,
+      "learning_rate": 9.145886369697618e-06,
+      "loss": 0.5452,
+      "step": 3346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7360591699213976,
+      "learning_rate": 9.145309639488835e-06,
+      "loss": 0.6373,
+      "step": 3347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5683055490433099,
+      "learning_rate": 9.144732732827464e-06,
+      "loss": 0.5132,
+      "step": 3348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8434869696891951,
+      "learning_rate": 9.144155649738067e-06,
+      "loss": 0.6544,
+      "step": 3349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6122219318901587,
+      "learning_rate": 9.143578390245207e-06,
+      "loss": 0.5217,
+      "step": 3350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9442240314101615,
+      "learning_rate": 9.143000954373455e-06,
+      "loss": 0.619,
+      "step": 3351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8609016654664363,
+      "learning_rate": 9.142423342147392e-06,
+      "loss": 0.6901,
+      "step": 3352
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8100428882519414,
+      "learning_rate": 9.141845553591606e-06,
+      "loss": 0.6494,
+      "step": 3353
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.761576052846754,
+      "learning_rate": 9.141267588730689e-06,
+      "loss": 0.6327,
+      "step": 3354
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6165467173058305,
+      "learning_rate": 9.140689447589245e-06,
+      "loss": 0.542,
+      "step": 3355
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5876969642271552,
+      "learning_rate": 9.140111130191882e-06,
+      "loss": 0.5399,
+      "step": 3356
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6287577389873411,
+      "learning_rate": 9.139532636563219e-06,
+      "loss": 0.5657,
+      "step": 3357
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8379049445511668,
+      "learning_rate": 9.13895396672788e-06,
+      "loss": 0.5946,
+      "step": 3358
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5379736932317782,
+      "learning_rate": 9.138375120710495e-06,
+      "loss": 0.5038,
+      "step": 3359
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8809318084173557,
+      "learning_rate": 9.137796098535707e-06,
+      "loss": 0.6579,
+      "step": 3360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6072689367456056,
+      "learning_rate": 9.137216900228161e-06,
+      "loss": 0.5149,
+      "step": 3361
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6422148248983705,
+      "learning_rate": 9.13663752581251e-06,
+      "loss": 0.5391,
+      "step": 3362
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6986428452209693,
+      "learning_rate": 9.136057975313422e-06,
+      "loss": 0.5689,
+      "step": 3363
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5706498515380872,
+      "learning_rate": 9.135478248755561e-06,
+      "loss": 0.5233,
+      "step": 3364
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6477492814285084,
+      "learning_rate": 9.134898346163608e-06,
+      "loss": 0.5261,
+      "step": 3365
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5451511723075446,
+      "learning_rate": 9.134318267562244e-06,
+      "loss": 0.4821,
+      "step": 3366
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0732824412215014,
+      "learning_rate": 9.133738012976163e-06,
+      "loss": 0.659,
+      "step": 3367
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6449010183592399,
+      "learning_rate": 9.133157582430063e-06,
+      "loss": 0.5152,
+      "step": 3368
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8101058622650507,
+      "learning_rate": 9.132576975948655e-06,
+      "loss": 0.5877,
+      "step": 3369
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7507506869939086,
+      "learning_rate": 9.13199619355665e-06,
+      "loss": 0.5631,
+      "step": 3370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.845321121448081,
+      "learning_rate": 9.13141523527877e-06,
+      "loss": 0.6395,
+      "step": 3371
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7082437538968397,
+      "learning_rate": 9.130834101139746e-06,
+      "loss": 0.5613,
+      "step": 3372
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9805342219127094,
+      "learning_rate": 9.130252791164316e-06,
+      "loss": 0.6344,
+      "step": 3373
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7900535866295054,
+      "learning_rate": 9.129671305377221e-06,
+      "loss": 0.6313,
+      "step": 3374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6552233602094087,
+      "learning_rate": 9.129089643803214e-06,
+      "loss": 0.629,
+      "step": 3375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6125297194096332,
+      "learning_rate": 9.12850780646706e-06,
+      "loss": 0.5558,
+      "step": 3376
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7039150157695548,
+      "learning_rate": 9.127925793393518e-06,
+      "loss": 0.5445,
+      "step": 3377
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5808338395472502,
+      "learning_rate": 9.127343604607368e-06,
+      "loss": 0.5222,
+      "step": 3378
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6737747923260493,
+      "learning_rate": 9.126761240133389e-06,
+      "loss": 0.5396,
+      "step": 3379
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7678252240305559,
+      "learning_rate": 9.126178699996372e-06,
+      "loss": 0.6594,
+      "step": 3380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6479400350903531,
+      "learning_rate": 9.125595984221111e-06,
+      "loss": 0.5199,
+      "step": 3381
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9702338280115503,
+      "learning_rate": 9.125013092832413e-06,
+      "loss": 0.6508,
+      "step": 3382
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7154203402310482,
+      "learning_rate": 9.124430025855092e-06,
+      "loss": 0.6366,
+      "step": 3383
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0165729780864998,
+      "learning_rate": 9.123846783313962e-06,
+      "loss": 0.6224,
+      "step": 3384
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5766769383579708,
+      "learning_rate": 9.123263365233853e-06,
+      "loss": 0.4986,
+      "step": 3385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6297472786360359,
+      "learning_rate": 9.122679771639597e-06,
+      "loss": 0.5538,
+      "step": 3386
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5713774395211625,
+      "learning_rate": 9.12209600255604e-06,
+      "loss": 0.4997,
+      "step": 3387
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7399484539715754,
+      "learning_rate": 9.121512058008027e-06,
+      "loss": 0.5827,
+      "step": 3388
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8690171513296493,
+      "learning_rate": 9.120927938020417e-06,
+      "loss": 0.5909,
+      "step": 3389
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6314486785053257,
+      "learning_rate": 9.120343642618072e-06,
+      "loss": 0.556,
+      "step": 3390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8256157442624091,
+      "learning_rate": 9.119759171825864e-06,
+      "loss": 0.5406,
+      "step": 3391
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3008687447765541,
+      "learning_rate": 9.119174525668675e-06,
+      "loss": 0.6182,
+      "step": 3392
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6699956479986262,
+      "learning_rate": 9.11858970417139e-06,
+      "loss": 0.5853,
+      "step": 3393
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.745843427690736,
+      "learning_rate": 9.1180047073589e-06,
+      "loss": 0.5609,
+      "step": 3394
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8526497815180284,
+      "learning_rate": 9.117419535256112e-06,
+      "loss": 0.6744,
+      "step": 3395
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8985836024702388,
+      "learning_rate": 9.116834187887929e-06,
+      "loss": 0.5944,
+      "step": 3396
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7569995747681864,
+      "learning_rate": 9.116248665279271e-06,
+      "loss": 0.5935,
+      "step": 3397
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8722201886654226,
+      "learning_rate": 9.115662967455062e-06,
+      "loss": 0.6522,
+      "step": 3398
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.54464972485816,
+      "learning_rate": 9.11507709444023e-06,
+      "loss": 0.4958,
+      "step": 3399
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9578844103435927,
+      "learning_rate": 9.11449104625972e-06,
+      "loss": 0.6634,
+      "step": 3400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6305311560784173,
+      "learning_rate": 9.11390482293847e-06,
+      "loss": 0.5234,
+      "step": 3401
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8473449763833505,
+      "learning_rate": 9.113318424501441e-06,
+      "loss": 0.6717,
+      "step": 3402
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6465775105355057,
+      "learning_rate": 9.112731850973588e-06,
+      "loss": 0.5333,
+      "step": 3403
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7232518515987104,
+      "learning_rate": 9.112145102379883e-06,
+      "loss": 0.6084,
+      "step": 3404
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.859278182864312,
+      "learning_rate": 9.111558178745303e-06,
+      "loss": 0.589,
+      "step": 3405
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0453940997207696,
+      "learning_rate": 9.110971080094831e-06,
+      "loss": 0.6317,
+      "step": 3406
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6580486255596497,
+      "learning_rate": 9.110383806453454e-06,
+      "loss": 0.531,
+      "step": 3407
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7649491544022339,
+      "learning_rate": 9.109796357846175e-06,
+      "loss": 0.648,
+      "step": 3408
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.759930269148066,
+      "learning_rate": 9.109208734297998e-06,
+      "loss": 0.6209,
+      "step": 3409
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7872410564615248,
+      "learning_rate": 9.108620935833936e-06,
+      "loss": 0.5486,
+      "step": 3410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9132832582965659,
+      "learning_rate": 9.10803296247901e-06,
+      "loss": 0.6446,
+      "step": 3411
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8071141709769595,
+      "learning_rate": 9.107444814258248e-06,
+      "loss": 0.6386,
+      "step": 3412
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6190393288077055,
+      "learning_rate": 9.106856491196687e-06,
+      "loss": 0.5587,
+      "step": 3413
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.687739849457782,
+      "learning_rate": 9.106267993319367e-06,
+      "loss": 0.5009,
+      "step": 3414
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7321852620574655,
+      "learning_rate": 9.105679320651342e-06,
+      "loss": 0.5614,
+      "step": 3415
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.795381282193837,
+      "learning_rate": 9.105090473217666e-06,
+      "loss": 0.5831,
+      "step": 3416
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5307703594494323,
+      "learning_rate": 9.104501451043408e-06,
+      "loss": 0.5021,
+      "step": 3417
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9021048770161111,
+      "learning_rate": 9.103912254153639e-06,
+      "loss": 0.5814,
+      "step": 3418
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.091798701074009,
+      "learning_rate": 9.10332288257344e-06,
+      "loss": 0.629,
+      "step": 3419
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7033812224130684,
+      "learning_rate": 9.102733336327898e-06,
+      "loss": 0.5506,
+      "step": 3420
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7485385640187735,
+      "learning_rate": 9.102143615442108e-06,
+      "loss": 0.5998,
+      "step": 3421
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8108791291501346,
+      "learning_rate": 9.101553719941174e-06,
+      "loss": 0.5812,
+      "step": 3422
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7281258194228255,
+      "learning_rate": 9.100963649850205e-06,
+      "loss": 0.6093,
+      "step": 3423
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.62751963414827,
+      "learning_rate": 9.100373405194315e-06,
+      "loss": 0.529,
+      "step": 3424
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.830029602085821,
+      "learning_rate": 9.099782985998636e-06,
+      "loss": 0.5853,
+      "step": 3425
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5672589507696492,
+      "learning_rate": 9.099192392288294e-06,
+      "loss": 0.5196,
+      "step": 3426
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8414099568616669,
+      "learning_rate": 9.098601624088431e-06,
+      "loss": 0.5821,
+      "step": 3427
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5549940027603152,
+      "learning_rate": 9.098010681424195e-06,
+      "loss": 0.5335,
+      "step": 3428
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5860203072757566,
+      "learning_rate": 9.09741956432074e-06,
+      "loss": 0.5245,
+      "step": 3429
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6877467215061256,
+      "learning_rate": 9.096828272803226e-06,
+      "loss": 0.5793,
+      "step": 3430
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7067325107093925,
+      "learning_rate": 9.096236806896826e-06,
+      "loss": 0.5951,
+      "step": 3431
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6761810453383481,
+      "learning_rate": 9.095645166626715e-06,
+      "loss": 0.5423,
+      "step": 3432
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5536052962497934,
+      "learning_rate": 9.095053352018075e-06,
+      "loss": 0.4714,
+      "step": 3433
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.781149231770683,
+      "learning_rate": 9.094461363096101e-06,
+      "loss": 0.6311,
+      "step": 3434
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8280277777453815,
+      "learning_rate": 9.093869199885989e-06,
+      "loss": 0.6185,
+      "step": 3435
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5771242618740802,
+      "learning_rate": 9.09327686241295e-06,
+      "loss": 0.4675,
+      "step": 3436
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.617704913127233,
+      "learning_rate": 9.092684350702194e-06,
+      "loss": 0.6161,
+      "step": 3437
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8166272347866761,
+      "learning_rate": 9.092091664778944e-06,
+      "loss": 0.6166,
+      "step": 3438
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6073872660581949,
+      "learning_rate": 9.091498804668427e-06,
+      "loss": 0.5854,
+      "step": 3439
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7709524855851002,
+      "learning_rate": 9.090905770395881e-06,
+      "loss": 0.6592,
+      "step": 3440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7446287029972817,
+      "learning_rate": 9.09031256198655e-06,
+      "loss": 0.5939,
+      "step": 3441
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5762656640439561,
+      "learning_rate": 9.089719179465684e-06,
+      "loss": 0.5321,
+      "step": 3442
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7922364264667471,
+      "learning_rate": 9.089125622858538e-06,
+      "loss": 0.6616,
+      "step": 3443
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5957511470102311,
+      "learning_rate": 9.088531892190387e-06,
+      "loss": 0.5604,
+      "step": 3444
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5609628154928852,
+      "learning_rate": 9.087937987486495e-06,
+      "loss": 0.4976,
+      "step": 3445
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6642831355527429,
+      "learning_rate": 9.087343908772147e-06,
+      "loss": 0.5502,
+      "step": 3446
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7742182344477593,
+      "learning_rate": 9.086749656072628e-06,
+      "loss": 0.6316,
+      "step": 3447
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7032856889729958,
+      "learning_rate": 9.086155229413237e-06,
+      "loss": 0.5739,
+      "step": 3448
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6828160605012141,
+      "learning_rate": 9.085560628819275e-06,
+      "loss": 0.6179,
+      "step": 3449
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6665142563842354,
+      "learning_rate": 9.084965854316053e-06,
+      "loss": 0.5718,
+      "step": 3450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6591496051120823,
+      "learning_rate": 9.084370905928888e-06,
+      "loss": 0.5358,
+      "step": 3451
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9074742145259302,
+      "learning_rate": 9.083775783683107e-06,
+      "loss": 0.6626,
+      "step": 3452
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7297895037645014,
+      "learning_rate": 9.083180487604037e-06,
+      "loss": 0.5436,
+      "step": 3453
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6099829711461648,
+      "learning_rate": 9.082585017717025e-06,
+      "loss": 0.5411,
+      "step": 3454
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.596012446359992,
+      "learning_rate": 9.081989374047414e-06,
+      "loss": 0.5547,
+      "step": 3455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6013586954300786,
+      "learning_rate": 9.081393556620559e-06,
+      "loss": 0.546,
+      "step": 3456
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6037784928475926,
+      "learning_rate": 9.080797565461823e-06,
+      "loss": 0.5251,
+      "step": 3457
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6740132365035242,
+      "learning_rate": 9.080201400596576e-06,
+      "loss": 0.5677,
+      "step": 3458
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5920250294100176,
+      "learning_rate": 9.079605062050192e-06,
+      "loss": 0.4839,
+      "step": 3459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8994227572042438,
+      "learning_rate": 9.079008549848059e-06,
+      "loss": 0.6469,
+      "step": 3460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7040440667533507,
+      "learning_rate": 9.078411864015564e-06,
+      "loss": 0.554,
+      "step": 3461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7005283401069226,
+      "learning_rate": 9.07781500457811e-06,
+      "loss": 0.5902,
+      "step": 3462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6336861352541178,
+      "learning_rate": 9.077217971561104e-06,
+      "loss": 0.5874,
+      "step": 3463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6950969304149969,
+      "learning_rate": 9.076620764989954e-06,
+      "loss": 0.5774,
+      "step": 3464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.663118890296144,
+      "learning_rate": 9.076023384890088e-06,
+      "loss": 0.5519,
+      "step": 3465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.740456861741743,
+      "learning_rate": 9.07542583128693e-06,
+      "loss": 0.6554,
+      "step": 3466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9448633073039158,
+      "learning_rate": 9.07482810420592e-06,
+      "loss": 0.6324,
+      "step": 3467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6265449999592396,
+      "learning_rate": 9.074230203672497e-06,
+      "loss": 0.5207,
+      "step": 3468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.773030939432317,
+      "learning_rate": 9.073632129712113e-06,
+      "loss": 0.6169,
+      "step": 3469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9168246701072386,
+      "learning_rate": 9.073033882350228e-06,
+      "loss": 0.6942,
+      "step": 3470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.531811412093806,
+      "learning_rate": 9.072435461612306e-06,
+      "loss": 0.633,
+      "step": 3471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6088111774191548,
+      "learning_rate": 9.07183686752382e-06,
+      "loss": 0.5624,
+      "step": 3472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5718113399475073,
+      "learning_rate": 9.07123810011025e-06,
+      "loss": 0.5179,
+      "step": 3473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6586860404694903,
+      "learning_rate": 9.070639159397083e-06,
+      "loss": 0.5209,
+      "step": 3474
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.695078600559415,
+      "learning_rate": 9.070040045409816e-06,
+      "loss": 0.543,
+      "step": 3475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7903743500897177,
+      "learning_rate": 9.06944075817395e-06,
+      "loss": 0.5196,
+      "step": 3476
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3566609457891983,
+      "learning_rate": 9.068841297714995e-06,
+      "loss": 0.6252,
+      "step": 3477
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8800272974540738,
+      "learning_rate": 9.068241664058468e-06,
+      "loss": 0.6801,
+      "step": 3478
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5566431404014974,
+      "learning_rate": 9.067641857229894e-06,
+      "loss": 0.5208,
+      "step": 3479
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6205159678325204,
+      "learning_rate": 9.067041877254804e-06,
+      "loss": 0.5347,
+      "step": 3480
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7549855744068191,
+      "learning_rate": 9.066441724158739e-06,
+      "loss": 0.5272,
+      "step": 3481
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5513535047703215,
+      "learning_rate": 9.065841397967244e-06,
+      "loss": 0.5126,
+      "step": 3482
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6202979237028544,
+      "learning_rate": 9.065240898705872e-06,
+      "loss": 0.5474,
+      "step": 3483
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.716092780306067,
+      "learning_rate": 9.064640226400185e-06,
+      "loss": 0.5148,
+      "step": 3484
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6400446680115223,
+      "learning_rate": 9.064039381075754e-06,
+      "loss": 0.5406,
+      "step": 3485
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.622689900777675,
+      "learning_rate": 9.063438362758152e-06,
+      "loss": 0.5448,
+      "step": 3486
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7307962940199746,
+      "learning_rate": 9.062837171472964e-06,
+      "loss": 0.5628,
+      "step": 3487
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6741559708179777,
+      "learning_rate": 9.062235807245782e-06,
+      "loss": 0.6137,
+      "step": 3488
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7451203092554143,
+      "learning_rate": 9.061634270102203e-06,
+      "loss": 0.6273,
+      "step": 3489
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7588593424679776,
+      "learning_rate": 9.06103256006783e-06,
+      "loss": 0.6269,
+      "step": 3490
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5981496309045811,
+      "learning_rate": 9.06043067716828e-06,
+      "loss": 0.5627,
+      "step": 3491
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6738724834755675,
+      "learning_rate": 9.05982862142917e-06,
+      "loss": 0.6255,
+      "step": 3492
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6156238216372788,
+      "learning_rate": 9.05922639287613e-06,
+      "loss": 0.5565,
+      "step": 3493
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6418613128208774,
+      "learning_rate": 9.058623991534792e-06,
+      "loss": 0.5136,
+      "step": 3494
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9065074820157594,
+      "learning_rate": 9.058021417430802e-06,
+      "loss": 0.6651,
+      "step": 3495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.877923466513029,
+      "learning_rate": 9.057418670589808e-06,
+      "loss": 0.6275,
+      "step": 3496
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6657402420828011,
+      "learning_rate": 9.056815751037467e-06,
+      "loss": 0.5542,
+      "step": 3497
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8721493810906396,
+      "learning_rate": 9.056212658799442e-06,
+      "loss": 0.6611,
+      "step": 3498
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6708404727325938,
+      "learning_rate": 9.055609393901406e-06,
+      "loss": 0.5685,
+      "step": 3499
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8400889025363306,
+      "learning_rate": 9.055005956369038e-06,
+      "loss": 0.6319,
+      "step": 3500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9255944785305048,
+      "learning_rate": 9.054402346228025e-06,
+      "loss": 0.6464,
+      "step": 3501
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6925266681120441,
+      "learning_rate": 9.05379856350406e-06,
+      "loss": 0.5551,
+      "step": 3502
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7483439418040335,
+      "learning_rate": 9.053194608222843e-06,
+      "loss": 0.5183,
+      "step": 3503
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6846736393015276,
+      "learning_rate": 9.052590480410086e-06,
+      "loss": 0.5299,
+      "step": 3504
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6688169423039391,
+      "learning_rate": 9.051986180091501e-06,
+      "loss": 0.6147,
+      "step": 3505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8562968352471371,
+      "learning_rate": 9.051381707292813e-06,
+      "loss": 0.6629,
+      "step": 3506
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5459875483422001,
+      "learning_rate": 9.050777062039753e-06,
+      "loss": 0.5054,
+      "step": 3507
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7191218558112576,
+      "learning_rate": 9.050172244358055e-06,
+      "loss": 0.6282,
+      "step": 3508
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9132885327842507,
+      "learning_rate": 9.04956725427347e-06,
+      "loss": 0.6207,
+      "step": 3509
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6305026284423193,
+      "learning_rate": 9.048962091811746e-06,
+      "loss": 0.5662,
+      "step": 3510
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.792983350898886,
+      "learning_rate": 9.048356756998645e-06,
+      "loss": 0.6244,
+      "step": 3511
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1881448784013917,
+      "learning_rate": 9.047751249859935e-06,
+      "loss": 0.6438,
+      "step": 3512
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7654548079755309,
+      "learning_rate": 9.047145570421387e-06,
+      "loss": 0.5731,
+      "step": 3513
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.882883690809048,
+      "learning_rate": 9.046539718708784e-06,
+      "loss": 0.6784,
+      "step": 3514
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6098511240638075,
+      "learning_rate": 9.045933694747919e-06,
+      "loss": 0.496,
+      "step": 3515
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8982423895289703,
+      "learning_rate": 9.045327498564584e-06,
+      "loss": 0.6329,
+      "step": 3516
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9235146585001086,
+      "learning_rate": 9.044721130184584e-06,
+      "loss": 0.5889,
+      "step": 3517
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6330000143716848,
+      "learning_rate": 9.04411458963373e-06,
+      "loss": 0.5864,
+      "step": 3518
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.764318534633108,
+      "learning_rate": 9.043507876937843e-06,
+      "loss": 0.5963,
+      "step": 3519
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8447809422758628,
+      "learning_rate": 9.042900992122745e-06,
+      "loss": 0.6525,
+      "step": 3520
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6936724517910646,
+      "learning_rate": 9.042293935214272e-06,
+      "loss": 0.4891,
+      "step": 3521
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6371401996807756,
+      "learning_rate": 9.041686706238264e-06,
+      "loss": 0.5634,
+      "step": 3522
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5988503216968053,
+      "learning_rate": 9.041079305220569e-06,
+      "loss": 0.5076,
+      "step": 3523
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9111324375489593,
+      "learning_rate": 9.04047173218704e-06,
+      "loss": 0.6961,
+      "step": 3524
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0821068528264417,
+      "learning_rate": 9.03986398716354e-06,
+      "loss": 0.6123,
+      "step": 3525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6841654508660304,
+      "learning_rate": 9.03925607017594e-06,
+      "loss": 0.5856,
+      "step": 3526
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8591711706905032,
+      "learning_rate": 9.038647981250117e-06,
+      "loss": 0.5892,
+      "step": 3527
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8684315769543314,
+      "learning_rate": 9.038039720411958e-06,
+      "loss": 0.5494,
+      "step": 3528
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1932004626529524,
+      "learning_rate": 9.037431287687348e-06,
+      "loss": 0.6754,
+      "step": 3529
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5471167506236436,
+      "learning_rate": 9.036822683102192e-06,
+      "loss": 0.5416,
+      "step": 3530
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3445573027619735,
+      "learning_rate": 9.036213906682392e-06,
+      "loss": 0.6121,
+      "step": 3531
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8696630568214259,
+      "learning_rate": 9.035604958453865e-06,
+      "loss": 0.7065,
+      "step": 3532
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.738741077272047,
+      "learning_rate": 9.034995838442533e-06,
+      "loss": 0.6173,
+      "step": 3533
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.683246772680504,
+      "learning_rate": 9.03438654667432e-06,
+      "loss": 0.6086,
+      "step": 3534
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6920271954019079,
+      "learning_rate": 9.033777083175165e-06,
+      "loss": 0.6018,
+      "step": 3535
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7884672789407592,
+      "learning_rate": 9.033167447971008e-06,
+      "loss": 0.6032,
+      "step": 3536
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5911953857692304,
+      "learning_rate": 9.0325576410878e-06,
+      "loss": 0.5322,
+      "step": 3537
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6238313942336,
+      "learning_rate": 9.031947662551502e-06,
+      "loss": 0.5774,
+      "step": 3538
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8525044260773812,
+      "learning_rate": 9.031337512388076e-06,
+      "loss": 0.6106,
+      "step": 3539
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7503718152639232,
+      "learning_rate": 9.030727190623495e-06,
+      "loss": 0.5538,
+      "step": 3540
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6340760337955164,
+      "learning_rate": 9.030116697283735e-06,
+      "loss": 0.5398,
+      "step": 3541
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7525091210116114,
+      "learning_rate": 9.029506032394787e-06,
+      "loss": 0.5367,
+      "step": 3542
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6724150312502559,
+      "learning_rate": 9.028895195982643e-06,
+      "loss": 0.5417,
+      "step": 3543
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.656939585729722,
+      "learning_rate": 9.028284188073306e-06,
+      "loss": 0.5082,
+      "step": 3544
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6931507465832119,
+      "learning_rate": 9.027673008692785e-06,
+      "loss": 0.605,
+      "step": 3545
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6786019314804285,
+      "learning_rate": 9.027061657867094e-06,
+      "loss": 0.5854,
+      "step": 3546
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7182869959058398,
+      "learning_rate": 9.026450135622254e-06,
+      "loss": 0.6483,
+      "step": 3547
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7704219718710181,
+      "learning_rate": 9.0258384419843e-06,
+      "loss": 0.601,
+      "step": 3548
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5832354117182584,
+      "learning_rate": 9.02522657697927e-06,
+      "loss": 0.539,
+      "step": 3549
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7130117431608659,
+      "learning_rate": 9.024614540633205e-06,
+      "loss": 0.5908,
+      "step": 3550
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6324806673749507,
+      "learning_rate": 9.02400233297216e-06,
+      "loss": 0.5251,
+      "step": 3551
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7298318214376163,
+      "learning_rate": 9.023389954022196e-06,
+      "loss": 0.6148,
+      "step": 3552
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5603837477940324,
+      "learning_rate": 9.02277740380938e-06,
+      "loss": 0.5157,
+      "step": 3553
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7181252072288951,
+      "learning_rate": 9.022164682359783e-06,
+      "loss": 0.5862,
+      "step": 3554
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2766420117783404,
+      "learning_rate": 9.021551789699487e-06,
+      "loss": 0.5757,
+      "step": 3555
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8044614780027008,
+      "learning_rate": 9.020938725854585e-06,
+      "loss": 0.6335,
+      "step": 3556
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6934871265053367,
+      "learning_rate": 9.020325490851169e-06,
+      "loss": 0.5562,
+      "step": 3557
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.565211036478897,
+      "learning_rate": 9.019712084715343e-06,
+      "loss": 0.5114,
+      "step": 3558
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7343693646958838,
+      "learning_rate": 9.019098507473222e-06,
+      "loss": 0.5775,
+      "step": 3559
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8307521926904984,
+      "learning_rate": 9.018484759150919e-06,
+      "loss": 0.6105,
+      "step": 3560
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6969391684108982,
+      "learning_rate": 9.017870839774561e-06,
+      "loss": 0.5618,
+      "step": 3561
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7913594337955373,
+      "learning_rate": 9.01725674937028e-06,
+      "loss": 0.6236,
+      "step": 3562
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7831484268905573,
+      "learning_rate": 9.016642487964216e-06,
+      "loss": 0.6271,
+      "step": 3563
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7734479276434101,
+      "learning_rate": 9.016028055582519e-06,
+      "loss": 0.5629,
+      "step": 3564
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9184675808772949,
+      "learning_rate": 9.015413452251339e-06,
+      "loss": 0.5913,
+      "step": 3565
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9242821490871466,
+      "learning_rate": 9.014798677996842e-06,
+      "loss": 0.6394,
+      "step": 3566
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5374049006998934,
+      "learning_rate": 9.014183732845192e-06,
+      "loss": 0.5085,
+      "step": 3567
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5652246223121693,
+      "learning_rate": 9.013568616822569e-06,
+      "loss": 0.5485,
+      "step": 3568
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.736919600344496,
+      "learning_rate": 9.012953329955155e-06,
+      "loss": 0.6425,
+      "step": 3569
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6557264578405426,
+      "learning_rate": 9.012337872269142e-06,
+      "loss": 0.4939,
+      "step": 3570
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6350277987555162,
+      "learning_rate": 9.011722243790728e-06,
+      "loss": 0.598,
+      "step": 3571
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8998325682651059,
+      "learning_rate": 9.011106444546115e-06,
+      "loss": 0.5951,
+      "step": 3572
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.854526313990166,
+      "learning_rate": 9.010490474561522e-06,
+      "loss": 0.6122,
+      "step": 3573
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7132548594570698,
+      "learning_rate": 9.009874333863163e-06,
+      "loss": 0.5614,
+      "step": 3574
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6080094417516144,
+      "learning_rate": 9.009258022477268e-06,
+      "loss": 0.4917,
+      "step": 3575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8485189554263428,
+      "learning_rate": 9.00864154043007e-06,
+      "loss": 0.6404,
+      "step": 3576
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8277463297538589,
+      "learning_rate": 9.008024887747813e-06,
+      "loss": 0.5755,
+      "step": 3577
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.967668426314493,
+      "learning_rate": 9.007408064456744e-06,
+      "loss": 0.6513,
+      "step": 3578
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6532989589050314,
+      "learning_rate": 9.006791070583119e-06,
+      "loss": 0.5663,
+      "step": 3579
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6288425525090104,
+      "learning_rate": 9.006173906153202e-06,
+      "loss": 0.4971,
+      "step": 3580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7285857204167445,
+      "learning_rate": 9.005556571193263e-06,
+      "loss": 0.5593,
+      "step": 3581
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6763956316419001,
+      "learning_rate": 9.004939065729581e-06,
+      "loss": 0.5647,
+      "step": 3582
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8032770269219566,
+      "learning_rate": 9.004321389788442e-06,
+      "loss": 0.6445,
+      "step": 3583
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.969687943895717,
+      "learning_rate": 9.003703543396137e-06,
+      "loss": 0.6297,
+      "step": 3584
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5827599786857293,
+      "learning_rate": 9.003085526578964e-06,
+      "loss": 0.5076,
+      "step": 3585
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6571114061701214,
+      "learning_rate": 9.002467339363234e-06,
+      "loss": 0.5265,
+      "step": 3586
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8832216698744784,
+      "learning_rate": 9.001848981775261e-06,
+      "loss": 0.5551,
+      "step": 3587
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7237677051629329,
+      "learning_rate": 9.001230453841363e-06,
+      "loss": 0.5705,
+      "step": 3588
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8385872730792924,
+      "learning_rate": 9.000611755587872e-06,
+      "loss": 0.6383,
+      "step": 3589
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8945242828773745,
+      "learning_rate": 8.99999288704112e-06,
+      "loss": 0.6061,
+      "step": 3590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7118769071711949,
+      "learning_rate": 8.999373848227455e-06,
+      "loss": 0.5276,
+      "step": 3591
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5566838614263709,
+      "learning_rate": 8.998754639173225e-06,
+      "loss": 0.5062,
+      "step": 3592
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6515824007018469,
+      "learning_rate": 8.998135259904787e-06,
+      "loss": 0.5533,
+      "step": 3593
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7030632564794325,
+      "learning_rate": 8.99751571044851e-06,
+      "loss": 0.5867,
+      "step": 3594
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7117488516645356,
+      "learning_rate": 8.996895990830763e-06,
+      "loss": 0.5591,
+      "step": 3595
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6918029758605921,
+      "learning_rate": 8.996276101077924e-06,
+      "loss": 0.5465,
+      "step": 3596
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6188471235386569,
+      "learning_rate": 8.995656041216383e-06,
+      "loss": 0.5254,
+      "step": 3597
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6665565780111213,
+      "learning_rate": 8.995035811272533e-06,
+      "loss": 0.5108,
+      "step": 3598
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6798739320660234,
+      "learning_rate": 8.994415411272772e-06,
+      "loss": 0.5485,
+      "step": 3599
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5883844255105386,
+      "learning_rate": 8.993794841243513e-06,
+      "loss": 0.4893,
+      "step": 3600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8426013484882545,
+      "learning_rate": 8.99317410121117e-06,
+      "loss": 0.6526,
+      "step": 3601
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9301523420229989,
+      "learning_rate": 8.992553191202168e-06,
+      "loss": 0.6579,
+      "step": 3602
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6248214648633588,
+      "learning_rate": 8.991932111242933e-06,
+      "loss": 0.5348,
+      "step": 3603
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8076042973562217,
+      "learning_rate": 8.991310861359904e-06,
+      "loss": 0.539,
+      "step": 3604
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7605421082766961,
+      "learning_rate": 8.990689441579526e-06,
+      "loss": 0.6825,
+      "step": 3605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7127814112166351,
+      "learning_rate": 8.990067851928252e-06,
+      "loss": 0.5872,
+      "step": 3606
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5811292756002087,
+      "learning_rate": 8.98944609243254e-06,
+      "loss": 0.5591,
+      "step": 3607
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5792800572504391,
+      "learning_rate": 8.988824163118854e-06,
+      "loss": 0.5058,
+      "step": 3608
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7089904525701487,
+      "learning_rate": 8.988202064013673e-06,
+      "loss": 0.5812,
+      "step": 3609
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7294877951057975,
+      "learning_rate": 8.987579795143473e-06,
+      "loss": 0.5835,
+      "step": 3610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6887990728814826,
+      "learning_rate": 8.986957356534742e-06,
+      "loss": 0.5641,
+      "step": 3611
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7191118812383622,
+      "learning_rate": 8.98633474821398e-06,
+      "loss": 0.5935,
+      "step": 3612
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6515297198485607,
+      "learning_rate": 8.985711970207685e-06,
+      "loss": 0.5361,
+      "step": 3613
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9157814976397483,
+      "learning_rate": 8.985089022542367e-06,
+      "loss": 0.6844,
+      "step": 3614
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7650958676349688,
+      "learning_rate": 8.984465905244546e-06,
+      "loss": 0.6407,
+      "step": 3615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.024815730251743,
+      "learning_rate": 8.983842618340742e-06,
+      "loss": 0.6333,
+      "step": 3616
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5460686408606807,
+      "learning_rate": 8.983219161857489e-06,
+      "loss": 0.5028,
+      "step": 3617
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6485308670297574,
+      "learning_rate": 8.982595535821324e-06,
+      "loss": 0.55,
+      "step": 3618
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6363369782361266,
+      "learning_rate": 8.981971740258795e-06,
+      "loss": 0.5457,
+      "step": 3619
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6064298546329495,
+      "learning_rate": 8.981347775196452e-06,
+      "loss": 0.572,
+      "step": 3620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.750851122173302,
+      "learning_rate": 8.980723640660857e-06,
+      "loss": 0.5652,
+      "step": 3621
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6175940383720777,
+      "learning_rate": 8.980099336678577e-06,
+      "loss": 0.5109,
+      "step": 3622
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6855898550667417,
+      "learning_rate": 8.979474863276188e-06,
+      "loss": 0.5377,
+      "step": 3623
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5832630637423732,
+      "learning_rate": 8.97885022048027e-06,
+      "loss": 0.5287,
+      "step": 3624
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6370558039230712,
+      "learning_rate": 8.978225408317414e-06,
+      "loss": 0.5674,
+      "step": 3625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7609008271334486,
+      "learning_rate": 8.977600426814215e-06,
+      "loss": 0.6438,
+      "step": 3626
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6273958182773944,
+      "learning_rate": 8.976975275997275e-06,
+      "loss": 0.5573,
+      "step": 3627
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9226487919657396,
+      "learning_rate": 8.976349955893206e-06,
+      "loss": 0.6231,
+      "step": 3628
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6128143406081149,
+      "learning_rate": 8.975724466528626e-06,
+      "loss": 0.5613,
+      "step": 3629
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6398142618401653,
+      "learning_rate": 8.97509880793016e-06,
+      "loss": 0.5611,
+      "step": 3630
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7264243852564435,
+      "learning_rate": 8.974472980124443e-06,
+      "loss": 0.5442,
+      "step": 3631
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5936566974062466,
+      "learning_rate": 8.973846983138107e-06,
+      "loss": 0.5127,
+      "step": 3632
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6381544609829135,
+      "learning_rate": 8.973220816997808e-06,
+      "loss": 0.5251,
+      "step": 3633
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6156416391308108,
+      "learning_rate": 8.972594481730192e-06,
+      "loss": 0.5123,
+      "step": 3634
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5261340639641087,
+      "learning_rate": 8.971967977361925e-06,
+      "loss": 0.5031,
+      "step": 3635
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.788753557283552,
+      "learning_rate": 8.971341303919676e-06,
+      "loss": 0.6076,
+      "step": 3636
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7121461437090074,
+      "learning_rate": 8.970714461430116e-06,
+      "loss": 0.5194,
+      "step": 3637
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6242108132600825,
+      "learning_rate": 8.970087449919929e-06,
+      "loss": 0.5454,
+      "step": 3638
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8317702218257098,
+      "learning_rate": 8.969460269415807e-06,
+      "loss": 0.6484,
+      "step": 3639
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6835746484984608,
+      "learning_rate": 8.968832919944446e-06,
+      "loss": 0.5384,
+      "step": 3640
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6736306798476158,
+      "learning_rate": 8.96820540153255e-06,
+      "loss": 0.5284,
+      "step": 3641
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0457220740733495,
+      "learning_rate": 8.967577714206832e-06,
+      "loss": 0.6925,
+      "step": 3642
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7113487787701108,
+      "learning_rate": 8.966949857994007e-06,
+      "loss": 0.6277,
+      "step": 3643
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7970221435192456,
+      "learning_rate": 8.966321832920804e-06,
+      "loss": 0.6078,
+      "step": 3644
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7886578393286339,
+      "learning_rate": 8.965693639013955e-06,
+      "loss": 0.5752,
+      "step": 3645
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7332633846795106,
+      "learning_rate": 8.965065276300202e-06,
+      "loss": 0.6483,
+      "step": 3646
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7502611305161161,
+      "learning_rate": 8.96443674480629e-06,
+      "loss": 0.5996,
+      "step": 3647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6608765133593425,
+      "learning_rate": 8.963808044558972e-06,
+      "loss": 0.5511,
+      "step": 3648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7846595181403333,
+      "learning_rate": 8.963179175585012e-06,
+      "loss": 0.5486,
+      "step": 3649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9753311951927206,
+      "learning_rate": 8.962550137911182e-06,
+      "loss": 0.5919,
+      "step": 3650
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6435971415344093,
+      "learning_rate": 8.961920931564255e-06,
+      "loss": 0.5739,
+      "step": 3651
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7819153363536666,
+      "learning_rate": 8.961291556571012e-06,
+      "loss": 0.644,
+      "step": 3652
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8680582180989188,
+      "learning_rate": 8.960662012958247e-06,
+      "loss": 0.6545,
+      "step": 3653
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.698502228725985,
+      "learning_rate": 8.960032300752756e-06,
+      "loss": 0.5272,
+      "step": 3654
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6257151794084105,
+      "learning_rate": 8.959402419981346e-06,
+      "loss": 0.5477,
+      "step": 3655
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.308083973214287,
+      "learning_rate": 8.958772370670826e-06,
+      "loss": 0.6393,
+      "step": 3656
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7685872453673079,
+      "learning_rate": 8.958142152848017e-06,
+      "loss": 0.6066,
+      "step": 3657
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7596884286710909,
+      "learning_rate": 8.957511766539745e-06,
+      "loss": 0.5981,
+      "step": 3658
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7544052857735296,
+      "learning_rate": 8.956881211772845e-06,
+      "loss": 0.6317,
+      "step": 3659
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8116571271042058,
+      "learning_rate": 8.956250488574155e-06,
+      "loss": 0.6562,
+      "step": 3660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8595715639648566,
+      "learning_rate": 8.955619596970523e-06,
+      "loss": 0.6507,
+      "step": 3661
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6162109935608634,
+      "learning_rate": 8.954988536988806e-06,
+      "loss": 0.5304,
+      "step": 3662
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7534583776145457,
+      "learning_rate": 8.954357308655866e-06,
+      "loss": 0.6389,
+      "step": 3663
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7576891010065256,
+      "learning_rate": 8.95372591199857e-06,
+      "loss": 0.6503,
+      "step": 3664
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5698501475385936,
+      "learning_rate": 8.953094347043797e-06,
+      "loss": 0.5285,
+      "step": 3665
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.695740224806512,
+      "learning_rate": 8.95246261381843e-06,
+      "loss": 0.5967,
+      "step": 3666
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6306413817937664,
+      "learning_rate": 8.95183071234936e-06,
+      "loss": 0.6064,
+      "step": 3667
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5847056946508924,
+      "learning_rate": 8.951198642663485e-06,
+      "loss": 0.5417,
+      "step": 3668
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7661917869993571,
+      "learning_rate": 8.95056640478771e-06,
+      "loss": 0.6035,
+      "step": 3669
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7672207861555538,
+      "learning_rate": 8.949933998748946e-06,
+      "loss": 0.5935,
+      "step": 3670
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7525470670952771,
+      "learning_rate": 8.949301424574115e-06,
+      "loss": 0.6315,
+      "step": 3671
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7143112124804997,
+      "learning_rate": 8.948668682290142e-06,
+      "loss": 0.5592,
+      "step": 3672
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7831403294970477,
+      "learning_rate": 8.948035771923961e-06,
+      "loss": 0.6157,
+      "step": 3673
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6399385467758807,
+      "learning_rate": 8.947402693502515e-06,
+      "loss": 0.4654,
+      "step": 3674
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8733134726545604,
+      "learning_rate": 8.94676944705275e-06,
+      "loss": 0.6258,
+      "step": 3675
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7460259667296016,
+      "learning_rate": 8.946136032601623e-06,
+      "loss": 0.5559,
+      "step": 3676
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8508096676925495,
+      "learning_rate": 8.945502450176092e-06,
+      "loss": 0.5815,
+      "step": 3677
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6754333686919946,
+      "learning_rate": 8.944868699803131e-06,
+      "loss": 0.5236,
+      "step": 3678
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5879271451746475,
+      "learning_rate": 8.944234781509715e-06,
+      "loss": 0.5497,
+      "step": 3679
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8066598560244146,
+      "learning_rate": 8.94360069532283e-06,
+      "loss": 0.5931,
+      "step": 3680
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.589504278045563,
+      "learning_rate": 8.942966441269464e-06,
+      "loss": 0.491,
+      "step": 3681
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5845327895512482,
+      "learning_rate": 8.942332019376616e-06,
+      "loss": 0.5332,
+      "step": 3682
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7246144716836019,
+      "learning_rate": 8.941697429671292e-06,
+      "loss": 0.6354,
+      "step": 3683
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5774564684260933,
+      "learning_rate": 8.941062672180504e-06,
+      "loss": 0.526,
+      "step": 3684
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.823466105597688,
+      "learning_rate": 8.940427746931273e-06,
+      "loss": 0.5982,
+      "step": 3685
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.261700497687932,
+      "learning_rate": 8.939792653950625e-06,
+      "loss": 0.6182,
+      "step": 3686
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6459124300892567,
+      "learning_rate": 8.939157393265595e-06,
+      "loss": 0.5445,
+      "step": 3687
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6322548075569477,
+      "learning_rate": 8.93852196490322e-06,
+      "loss": 0.5482,
+      "step": 3688
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.797144074997912,
+      "learning_rate": 8.93788636889055e-06,
+      "loss": 0.5824,
+      "step": 3689
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.746353141671459,
+      "learning_rate": 8.937250605254641e-06,
+      "loss": 0.5989,
+      "step": 3690
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8253001390495809,
+      "learning_rate": 8.936614674022556e-06,
+      "loss": 0.6158,
+      "step": 3691
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8076365924064134,
+      "learning_rate": 8.935978575221366e-06,
+      "loss": 0.6325,
+      "step": 3692
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5585774221205613,
+      "learning_rate": 8.935342308878144e-06,
+      "loss": 0.4789,
+      "step": 3693
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6422880908803965,
+      "learning_rate": 8.934705875019976e-06,
+      "loss": 0.563,
+      "step": 3694
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9356429250556946,
+      "learning_rate": 8.934069273673951e-06,
+      "loss": 0.5969,
+      "step": 3695
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6924930173721241,
+      "learning_rate": 8.93343250486717e-06,
+      "loss": 0.5042,
+      "step": 3696
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6793333757864841,
+      "learning_rate": 8.932795568626736e-06,
+      "loss": 0.4747,
+      "step": 3697
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6395061693133614,
+      "learning_rate": 8.93215846497976e-06,
+      "loss": 0.5652,
+      "step": 3698
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.789979768524907,
+      "learning_rate": 8.931521193953367e-06,
+      "loss": 0.573,
+      "step": 3699
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9383309760443929,
+      "learning_rate": 8.930883755574678e-06,
+      "loss": 0.6093,
+      "step": 3700
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7544976213925937,
+      "learning_rate": 8.93024614987083e-06,
+      "loss": 0.676,
+      "step": 3701
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7721047466971459,
+      "learning_rate": 8.929608376868963e-06,
+      "loss": 0.63,
+      "step": 3702
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5767679134176217,
+      "learning_rate": 8.928970436596223e-06,
+      "loss": 0.4625,
+      "step": 3703
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7562580186526595,
+      "learning_rate": 8.928332329079767e-06,
+      "loss": 0.6102,
+      "step": 3704
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.588729134635355,
+      "learning_rate": 8.927694054346758e-06,
+      "loss": 0.5789,
+      "step": 3705
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7680495111791352,
+      "learning_rate": 8.927055612424365e-06,
+      "loss": 0.6219,
+      "step": 3706
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6400531904276301,
+      "learning_rate": 8.926417003339762e-06,
+      "loss": 0.5259,
+      "step": 3707
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6218218356598285,
+      "learning_rate": 8.925778227120135e-06,
+      "loss": 0.5129,
+      "step": 3708
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7326078904729264,
+      "learning_rate": 8.925139283792672e-06,
+      "loss": 0.6119,
+      "step": 3709
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6507012789068991,
+      "learning_rate": 8.924500173384575e-06,
+      "loss": 0.5466,
+      "step": 3710
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8024968920410219,
+      "learning_rate": 8.923860895923045e-06,
+      "loss": 0.6062,
+      "step": 3711
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6992840744481165,
+      "learning_rate": 8.923221451435297e-06,
+      "loss": 0.565,
+      "step": 3712
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6309224522411508,
+      "learning_rate": 8.922581839948548e-06,
+      "loss": 0.5373,
+      "step": 3713
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7646701275992417,
+      "learning_rate": 8.921942061490023e-06,
+      "loss": 0.6289,
+      "step": 3714
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7521383474272808,
+      "learning_rate": 8.92130211608696e-06,
+      "loss": 0.6059,
+      "step": 3715
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8684102885098306,
+      "learning_rate": 8.920662003766595e-06,
+      "loss": 0.6571,
+      "step": 3716
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7495562426728156,
+      "learning_rate": 8.920021724556178e-06,
+      "loss": 0.6,
+      "step": 3717
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6088179970275477,
+      "learning_rate": 8.919381278482962e-06,
+      "loss": 0.5447,
+      "step": 3718
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8082923425081123,
+      "learning_rate": 8.91874066557421e-06,
+      "loss": 0.6528,
+      "step": 3719
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5844725058247551,
+      "learning_rate": 8.91809988585719e-06,
+      "loss": 0.5033,
+      "step": 3720
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9171774230903313,
+      "learning_rate": 8.917458939359178e-06,
+      "loss": 0.648,
+      "step": 3721
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6941438210718455,
+      "learning_rate": 8.91681782610746e-06,
+      "loss": 0.5536,
+      "step": 3722
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6281803328648323,
+      "learning_rate": 8.91617654612932e-06,
+      "loss": 0.4978,
+      "step": 3723
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6482814929124665,
+      "learning_rate": 8.915535099452061e-06,
+      "loss": 0.538,
+      "step": 3724
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7974226398465755,
+      "learning_rate": 8.914893486102983e-06,
+      "loss": 0.5559,
+      "step": 3725
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6514447406437415,
+      "learning_rate": 8.914251706109402e-06,
+      "loss": 0.5573,
+      "step": 3726
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8363222865577461,
+      "learning_rate": 8.913609759498631e-06,
+      "loss": 0.6463,
+      "step": 3727
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9874203579039861,
+      "learning_rate": 8.912967646298001e-06,
+      "loss": 0.6479,
+      "step": 3728
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8192879669837126,
+      "learning_rate": 8.912325366534842e-06,
+      "loss": 0.6297,
+      "step": 3729
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5820980775361936,
+      "learning_rate": 8.911682920236495e-06,
+      "loss": 0.5041,
+      "step": 3730
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5925154836169583,
+      "learning_rate": 8.911040307430303e-06,
+      "loss": 0.5716,
+      "step": 3731
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9268843463803044,
+      "learning_rate": 8.910397528143627e-06,
+      "loss": 0.6361,
+      "step": 3732
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8574553628574446,
+      "learning_rate": 8.909754582403822e-06,
+      "loss": 0.6125,
+      "step": 3733
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0145060814788536,
+      "learning_rate": 8.909111470238256e-06,
+      "loss": 0.6787,
+      "step": 3734
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5704184117156543,
+      "learning_rate": 8.90846819167431e-06,
+      "loss": 0.5245,
+      "step": 3735
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3110162758325523,
+      "learning_rate": 8.907824746739363e-06,
+      "loss": 0.6501,
+      "step": 3736
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9043482548097569,
+      "learning_rate": 8.9071811354608e-06,
+      "loss": 0.6681,
+      "step": 3737
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.709447598432559,
+      "learning_rate": 8.906537357866026e-06,
+      "loss": 0.5274,
+      "step": 3738
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0643688626522934,
+      "learning_rate": 8.905893413982438e-06,
+      "loss": 0.6129,
+      "step": 3739
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6010955615352654,
+      "learning_rate": 8.905249303837448e-06,
+      "loss": 0.588,
+      "step": 3740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6190471340716149,
+      "learning_rate": 8.904605027458477e-06,
+      "loss": 0.5585,
+      "step": 3741
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8350867052162392,
+      "learning_rate": 8.903960584872945e-06,
+      "loss": 0.653,
+      "step": 3742
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6357271986644341,
+      "learning_rate": 8.90331597610829e-06,
+      "loss": 0.538,
+      "step": 3743
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9068142955416113,
+      "learning_rate": 8.90267120119194e-06,
+      "loss": 0.6197,
+      "step": 3744
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6219985719635462,
+      "learning_rate": 8.902026260151355e-06,
+      "loss": 0.5764,
+      "step": 3745
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7068264692512304,
+      "learning_rate": 8.90138115301398e-06,
+      "loss": 0.5794,
+      "step": 3746
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7212069131587737,
+      "learning_rate": 8.900735879807274e-06,
+      "loss": 0.5856,
+      "step": 3747
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8011842512755848,
+      "learning_rate": 8.900090440558706e-06,
+      "loss": 0.6483,
+      "step": 3748
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7802673096194498,
+      "learning_rate": 8.899444835295754e-06,
+      "loss": 0.6121,
+      "step": 3749
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8161764962167143,
+      "learning_rate": 8.898799064045895e-06,
+      "loss": 0.6432,
+      "step": 3750
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7955542280150687,
+      "learning_rate": 8.898153126836618e-06,
+      "loss": 0.6138,
+      "step": 3751
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6364499303774556,
+      "learning_rate": 8.897507023695418e-06,
+      "loss": 0.5281,
+      "step": 3752
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.819956479796958,
+      "learning_rate": 8.896860754649801e-06,
+      "loss": 0.65,
+      "step": 3753
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6882934163630536,
+      "learning_rate": 8.896214319727273e-06,
+      "loss": 0.611,
+      "step": 3754
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.71638994144973,
+      "learning_rate": 8.895567718955353e-06,
+      "loss": 0.5825,
+      "step": 3755
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.722479875569419,
+      "learning_rate": 8.894920952361562e-06,
+      "loss": 0.6253,
+      "step": 3756
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1841776943865747,
+      "learning_rate": 8.894274019973433e-06,
+      "loss": 0.6373,
+      "step": 3757
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7555028605921302,
+      "learning_rate": 8.893626921818504e-06,
+      "loss": 0.5653,
+      "step": 3758
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7476453365424197,
+      "learning_rate": 8.89297965792432e-06,
+      "loss": 0.6065,
+      "step": 3759
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5876891237970107,
+      "learning_rate": 8.89233222831843e-06,
+      "loss": 0.5155,
+      "step": 3760
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5492884386908883,
+      "learning_rate": 8.891684633028397e-06,
+      "loss": 0.4743,
+      "step": 3761
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6953441699789937,
+      "learning_rate": 8.891036872081784e-06,
+      "loss": 0.5283,
+      "step": 3762
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5973827674200589,
+      "learning_rate": 8.890388945506166e-06,
+      "loss": 0.5016,
+      "step": 3763
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7751216169597506,
+      "learning_rate": 8.889740853329121e-06,
+      "loss": 0.5599,
+      "step": 3764
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8330327571955368,
+      "learning_rate": 8.88909259557824e-06,
+      "loss": 0.5825,
+      "step": 3765
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7576639230977771,
+      "learning_rate": 8.888444172281112e-06,
+      "loss": 0.6222,
+      "step": 3766
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.639892378999344,
+      "learning_rate": 8.887795583465343e-06,
+      "loss": 0.5417,
+      "step": 3767
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7382001861606764,
+      "learning_rate": 8.88714682915854e-06,
+      "loss": 0.6214,
+      "step": 3768
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8397368465145428,
+      "learning_rate": 8.886497909388318e-06,
+      "loss": 0.697,
+      "step": 3769
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6448698628371455,
+      "learning_rate": 8.885848824182299e-06,
+      "loss": 0.5623,
+      "step": 3770
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5877626713984844,
+      "learning_rate": 8.885199573568113e-06,
+      "loss": 0.5625,
+      "step": 3771
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7280144982047488,
+      "learning_rate": 8.884550157573398e-06,
+      "loss": 0.5979,
+      "step": 3772
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.846620867906846,
+      "learning_rate": 8.883900576225795e-06,
+      "loss": 0.6208,
+      "step": 3773
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9117724489057908,
+      "learning_rate": 8.883250829552954e-06,
+      "loss": 0.5988,
+      "step": 3774
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6069774680218666,
+      "learning_rate": 8.882600917582535e-06,
+      "loss": 0.4871,
+      "step": 3775
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6308346196193754,
+      "learning_rate": 8.881950840342204e-06,
+      "loss": 0.5341,
+      "step": 3776
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.692961978011267,
+      "learning_rate": 8.881300597859628e-06,
+      "loss": 0.5472,
+      "step": 3777
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6725794310638088,
+      "learning_rate": 8.88065019016249e-06,
+      "loss": 0.5911,
+      "step": 3778
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7646460223700365,
+      "learning_rate": 8.879999617278474e-06,
+      "loss": 0.5878,
+      "step": 3779
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3992359439150255,
+      "learning_rate": 8.879348879235273e-06,
+      "loss": 0.6499,
+      "step": 3780
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5100420033806663,
+      "learning_rate": 8.878697976060587e-06,
+      "loss": 0.6436,
+      "step": 3781
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6949835926994373,
+      "learning_rate": 8.878046907782123e-06,
+      "loss": 0.5458,
+      "step": 3782
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7426340637441344,
+      "learning_rate": 8.877395674427594e-06,
+      "loss": 0.5895,
+      "step": 3783
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9767401982713976,
+      "learning_rate": 8.87674427602472e-06,
+      "loss": 0.5982,
+      "step": 3784
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.739690809890258,
+      "learning_rate": 8.876092712601232e-06,
+      "loss": 0.6252,
+      "step": 3785
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.535623539253013,
+      "learning_rate": 8.875440984184865e-06,
+      "loss": 0.5018,
+      "step": 3786
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6071631986355673,
+      "learning_rate": 8.874789090803357e-06,
+      "loss": 0.5932,
+      "step": 3787
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.743789965590836,
+      "learning_rate": 8.87413703248446e-06,
+      "loss": 0.6023,
+      "step": 3788
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6378897141467588,
+      "learning_rate": 8.87348480925593e-06,
+      "loss": 0.5606,
+      "step": 3789
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7544153680391198,
+      "learning_rate": 8.87283242114553e-06,
+      "loss": 0.6267,
+      "step": 3790
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.694397722754607,
+      "learning_rate": 8.872179868181028e-06,
+      "loss": 0.6275,
+      "step": 3791
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8056110231682484,
+      "learning_rate": 8.871527150390205e-06,
+      "loss": 0.5115,
+      "step": 3792
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7209354821466226,
+      "learning_rate": 8.870874267800839e-06,
+      "loss": 0.5651,
+      "step": 3793
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6068038272840759,
+      "learning_rate": 8.870221220440729e-06,
+      "loss": 0.5354,
+      "step": 3794
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6011028532546561,
+      "learning_rate": 8.869568008337669e-06,
+      "loss": 0.5603,
+      "step": 3795
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0632137333295957,
+      "learning_rate": 8.86891463151946e-06,
+      "loss": 0.6775,
+      "step": 3796
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6163072991381189,
+      "learning_rate": 8.868261090013923e-06,
+      "loss": 0.5296,
+      "step": 3797
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6114030165969825,
+      "learning_rate": 8.86760738384887e-06,
+      "loss": 0.5308,
+      "step": 3798
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5891162191896753,
+      "learning_rate": 8.86695351305213e-06,
+      "loss": 0.5534,
+      "step": 3799
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7547945331350864,
+      "learning_rate": 8.866299477651538e-06,
+      "loss": 0.5845,
+      "step": 3800
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5839010840818549,
+      "learning_rate": 8.86564527767493e-06,
+      "loss": 0.4936,
+      "step": 3801
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.851915511901835,
+      "learning_rate": 8.864990913150157e-06,
+      "loss": 0.6214,
+      "step": 3802
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6491400211590191,
+      "learning_rate": 8.86433638410507e-06,
+      "loss": 0.5591,
+      "step": 3803
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6449044857682525,
+      "learning_rate": 8.863681690567533e-06,
+      "loss": 0.5224,
+      "step": 3804
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6280119347772116,
+      "learning_rate": 8.863026832565412e-06,
+      "loss": 0.5718,
+      "step": 3805
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.594692193545,
+      "learning_rate": 8.862371810126584e-06,
+      "loss": 0.5003,
+      "step": 3806
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6855556235983449,
+      "learning_rate": 8.86171662327893e-06,
+      "loss": 0.5872,
+      "step": 3807
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7636411056548783,
+      "learning_rate": 8.861061272050339e-06,
+      "loss": 0.6026,
+      "step": 3808
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6804317876899881,
+      "learning_rate": 8.860405756468709e-06,
+      "loss": 0.5311,
+      "step": 3809
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.756783970454099,
+      "learning_rate": 8.859750076561942e-06,
+      "loss": 0.6205,
+      "step": 3810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7552420388713613,
+      "learning_rate": 8.859094232357948e-06,
+      "loss": 0.6087,
+      "step": 3811
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8209982992660904,
+      "learning_rate": 8.858438223884644e-06,
+      "loss": 0.6055,
+      "step": 3812
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6056568586874923,
+      "learning_rate": 8.857782051169956e-06,
+      "loss": 0.5688,
+      "step": 3813
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8394050062298642,
+      "learning_rate": 8.857125714241812e-06,
+      "loss": 0.6111,
+      "step": 3814
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6203192329381239,
+      "learning_rate": 8.856469213128153e-06,
+      "loss": 0.5345,
+      "step": 3815
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.694377448647183,
+      "learning_rate": 8.855812547856923e-06,
+      "loss": 0.5889,
+      "step": 3816
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8583079935692607,
+      "learning_rate": 8.855155718456075e-06,
+      "loss": 0.6116,
+      "step": 3817
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7022409878542383,
+      "learning_rate": 8.854498724953564e-06,
+      "loss": 0.6024,
+      "step": 3818
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7305438770594196,
+      "learning_rate": 8.853841567377363e-06,
+      "loss": 0.5816,
+      "step": 3819
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6651848706841397,
+      "learning_rate": 8.853184245755439e-06,
+      "loss": 0.5932,
+      "step": 3820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8874623586818131,
+      "learning_rate": 8.852526760115778e-06,
+      "loss": 0.6175,
+      "step": 3821
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6640870272319327,
+      "learning_rate": 8.851869110486362e-06,
+      "loss": 0.534,
+      "step": 3822
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5720754296337881,
+      "learning_rate": 8.851211296895186e-06,
+      "loss": 0.4845,
+      "step": 3823
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8111147186094624,
+      "learning_rate": 8.850553319370252e-06,
+      "loss": 0.6294,
+      "step": 3824
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7863860785012166,
+      "learning_rate": 8.84989517793957e-06,
+      "loss": 0.6354,
+      "step": 3825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8385381607676841,
+      "learning_rate": 8.84923687263115e-06,
+      "loss": 0.6259,
+      "step": 3826
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7224318901767393,
+      "learning_rate": 8.848578403473018e-06,
+      "loss": 0.5923,
+      "step": 3827
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8064197154174032,
+      "learning_rate": 8.847919770493201e-06,
+      "loss": 0.6581,
+      "step": 3828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.570652831295744,
+      "learning_rate": 8.847260973719736e-06,
+      "loss": 0.562,
+      "step": 3829
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7335012936634536,
+      "learning_rate": 8.846602013180666e-06,
+      "loss": 0.5797,
+      "step": 3830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6228570584062498,
+      "learning_rate": 8.845942888904038e-06,
+      "loss": 0.5717,
+      "step": 3831
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7002958417312181,
+      "learning_rate": 8.845283600917914e-06,
+      "loss": 0.5549,
+      "step": 3832
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.636369277306478,
+      "learning_rate": 8.844624149250354e-06,
+      "loss": 0.6098,
+      "step": 3833
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7456172052339775,
+      "learning_rate": 8.84396453392943e-06,
+      "loss": 0.5924,
+      "step": 3834
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9141506628012072,
+      "learning_rate": 8.843304754983217e-06,
+      "loss": 0.6215,
+      "step": 3835
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9676529684471039,
+      "learning_rate": 8.842644812439804e-06,
+      "loss": 0.6414,
+      "step": 3836
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7605335745927064,
+      "learning_rate": 8.84198470632728e-06,
+      "loss": 0.6494,
+      "step": 3837
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7200340824015131,
+      "learning_rate": 8.841324436673745e-06,
+      "loss": 0.6169,
+      "step": 3838
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6287154856209565,
+      "learning_rate": 8.840664003507304e-06,
+      "loss": 0.5316,
+      "step": 3839
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8146697148983737,
+      "learning_rate": 8.840003406856067e-06,
+      "loss": 0.5974,
+      "step": 3840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7181890454293451,
+      "learning_rate": 8.83934264674816e-06,
+      "loss": 0.5717,
+      "step": 3841
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6613350860317138,
+      "learning_rate": 8.838681723211701e-06,
+      "loss": 0.5262,
+      "step": 3842
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.791346358642499,
+      "learning_rate": 8.838020636274832e-06,
+      "loss": 0.6149,
+      "step": 3843
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7061119137167481,
+      "learning_rate": 8.837359385965688e-06,
+      "loss": 0.5473,
+      "step": 3844
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8103306329770703,
+      "learning_rate": 8.836697972312413e-06,
+      "loss": 0.6177,
+      "step": 3845
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.765865468795884,
+      "learning_rate": 8.83603639534317e-06,
+      "loss": 0.6611,
+      "step": 3846
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1176111088317007,
+      "learning_rate": 8.835374655086116e-06,
+      "loss": 0.5709,
+      "step": 3847
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5733228264454907,
+      "learning_rate": 8.834712751569417e-06,
+      "loss": 0.4513,
+      "step": 3848
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8420986077281761,
+      "learning_rate": 8.834050684821251e-06,
+      "loss": 0.5911,
+      "step": 3849
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.615357020898463,
+      "learning_rate": 8.8333884548698e-06,
+      "loss": 0.5186,
+      "step": 3850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6185145452469081,
+      "learning_rate": 8.83272606174325e-06,
+      "loss": 0.5413,
+      "step": 3851
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7200094600005802,
+      "learning_rate": 8.832063505469801e-06,
+      "loss": 0.5959,
+      "step": 3852
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5578276948046608,
+      "learning_rate": 8.831400786077653e-06,
+      "loss": 0.5372,
+      "step": 3853
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5840033030101617,
+      "learning_rate": 8.830737903595019e-06,
+      "loss": 0.5608,
+      "step": 3854
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8529316521621446,
+      "learning_rate": 8.830074858050111e-06,
+      "loss": 0.6426,
+      "step": 3855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5460277659603923,
+      "learning_rate": 8.829411649471158e-06,
+      "loss": 0.5459,
+      "step": 3856
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6497799968015209,
+      "learning_rate": 8.828748277886386e-06,
+      "loss": 0.5793,
+      "step": 3857
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6160836682642195,
+      "learning_rate": 8.828084743324036e-06,
+      "loss": 0.5312,
+      "step": 3858
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5674052953999604,
+      "learning_rate": 8.82742104581235e-06,
+      "loss": 0.506,
+      "step": 3859
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6026918906289411,
+      "learning_rate": 8.826757185379582e-06,
+      "loss": 0.5994,
+      "step": 3860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6268201867888585,
+      "learning_rate": 8.826093162053989e-06,
+      "loss": 0.4832,
+      "step": 3861
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7433241407554696,
+      "learning_rate": 8.825428975863836e-06,
+      "loss": 0.5958,
+      "step": 3862
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.78294957505671,
+      "learning_rate": 8.824764626837397e-06,
+      "loss": 0.6129,
+      "step": 3863
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7157255879477162,
+      "learning_rate": 8.82410011500295e-06,
+      "loss": 0.5746,
+      "step": 3864
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6111326643348116,
+      "learning_rate": 8.823435440388779e-06,
+      "loss": 0.475,
+      "step": 3865
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6872514168090972,
+      "learning_rate": 8.822770603023181e-06,
+      "loss": 0.5695,
+      "step": 3866
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8509162722683828,
+      "learning_rate": 8.822105602934454e-06,
+      "loss": 0.6404,
+      "step": 3867
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6578410806098519,
+      "learning_rate": 8.821440440150906e-06,
+      "loss": 0.5242,
+      "step": 3868
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6987518175746289,
+      "learning_rate": 8.82077511470085e-06,
+      "loss": 0.5508,
+      "step": 3869
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0894242529125457,
+      "learning_rate": 8.820109626612604e-06,
+      "loss": 0.6254,
+      "step": 3870
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.853879069248099,
+      "learning_rate": 8.8194439759145e-06,
+      "loss": 0.6686,
+      "step": 3871
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8424280450962335,
+      "learning_rate": 8.818778162634874e-06,
+      "loss": 0.6449,
+      "step": 3872
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8413090703550339,
+      "learning_rate": 8.818112186802062e-06,
+      "loss": 0.6777,
+      "step": 3873
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6179861901530256,
+      "learning_rate": 8.817446048444415e-06,
+      "loss": 0.5722,
+      "step": 3874
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6634789003231849,
+      "learning_rate": 8.816779747590289e-06,
+      "loss": 0.5636,
+      "step": 3875
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7289076672701938,
+      "learning_rate": 8.816113284268046e-06,
+      "loss": 0.5762,
+      "step": 3876
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6968440077589,
+      "learning_rate": 8.815446658506056e-06,
+      "loss": 0.5641,
+      "step": 3877
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5641354164248852,
+      "learning_rate": 8.814779870332692e-06,
+      "loss": 0.5449,
+      "step": 3878
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6193433571350114,
+      "learning_rate": 8.81411291977634e-06,
+      "loss": 0.5039,
+      "step": 3879
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8229145202184409,
+      "learning_rate": 8.81344580686539e-06,
+      "loss": 0.637,
+      "step": 3880
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8346751370246499,
+      "learning_rate": 8.812778531628238e-06,
+      "loss": 0.6622,
+      "step": 3881
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1992556023933005,
+      "learning_rate": 8.812111094093287e-06,
+      "loss": 0.6383,
+      "step": 3882
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5229181416011706,
+      "learning_rate": 8.811443494288949e-06,
+      "loss": 0.5322,
+      "step": 3883
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5361558155327227,
+      "learning_rate": 8.81077573224364e-06,
+      "loss": 0.5464,
+      "step": 3884
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.680848544029737,
+      "learning_rate": 8.810107807985787e-06,
+      "loss": 0.6104,
+      "step": 3885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7053434025936305,
+      "learning_rate": 8.80943972154382e-06,
+      "loss": 0.5591,
+      "step": 3886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.763889628092451,
+      "learning_rate": 8.808771472946176e-06,
+      "loss": 0.5878,
+      "step": 3887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7282180768583473,
+      "learning_rate": 8.808103062221303e-06,
+      "loss": 0.5763,
+      "step": 3888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8156709587893172,
+      "learning_rate": 8.80743448939765e-06,
+      "loss": 0.6678,
+      "step": 3889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6750448945648101,
+      "learning_rate": 8.806765754503679e-06,
+      "loss": 0.568,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6642955876761903,
+      "learning_rate": 8.806096857567854e-06,
+      "loss": 0.5051,
+      "step": 3891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7798986842305553,
+      "learning_rate": 8.805427798618647e-06,
+      "loss": 0.6624,
+      "step": 3892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5693023559753447,
+      "learning_rate": 8.804758577684541e-06,
+      "loss": 0.5417,
+      "step": 3893
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7554695739250997,
+      "learning_rate": 8.80408919479402e-06,
+      "loss": 0.6065,
+      "step": 3894
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7597579230078431,
+      "learning_rate": 8.803419649975577e-06,
+      "loss": 0.5305,
+      "step": 3895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6083369015618603,
+      "learning_rate": 8.802749943257715e-06,
+      "loss": 0.5535,
+      "step": 3896
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6535379816174423,
+      "learning_rate": 8.80208007466894e-06,
+      "loss": 0.6076,
+      "step": 3897
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6099357229501594,
+      "learning_rate": 8.801410044237765e-06,
+      "loss": 0.5075,
+      "step": 3898
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8653682857511228,
+      "learning_rate": 8.800739851992712e-06,
+      "loss": 0.6629,
+      "step": 3899
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7785614415931186,
+      "learning_rate": 8.800069497962309e-06,
+      "loss": 0.5511,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8210474196867106,
+      "learning_rate": 8.799398982175091e-06,
+      "loss": 0.6493,
+      "step": 3901
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8364813898871233,
+      "learning_rate": 8.7987283046596e-06,
+      "loss": 0.6044,
+      "step": 3902
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6456465765628633,
+      "learning_rate": 8.798057465444383e-06,
+      "loss": 0.5518,
+      "step": 3903
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6409540453409056,
+      "learning_rate": 8.797386464557998e-06,
+      "loss": 0.5093,
+      "step": 3904
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7971425482188506,
+      "learning_rate": 8.796715302029004e-06,
+      "loss": 0.6072,
+      "step": 3905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8016859291483994,
+      "learning_rate": 8.796043977885973e-06,
+      "loss": 0.5869,
+      "step": 3906
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8663166529864562,
+      "learning_rate": 8.79537249215748e-06,
+      "loss": 0.6501,
+      "step": 3907
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6453729917214116,
+      "learning_rate": 8.794700844872108e-06,
+      "loss": 0.5609,
+      "step": 3908
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6642357020406892,
+      "learning_rate": 8.794029036058449e-06,
+      "loss": 0.5729,
+      "step": 3909
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6688130520578079,
+      "learning_rate": 8.793357065745097e-06,
+      "loss": 0.6118,
+      "step": 3910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8056793290954218,
+      "learning_rate": 8.792684933960656e-06,
+      "loss": 0.6105,
+      "step": 3911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.793202100360878,
+      "learning_rate": 8.792012640733735e-06,
+      "loss": 0.6141,
+      "step": 3912
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6599718294437324,
+      "learning_rate": 8.791340186092958e-06,
+      "loss": 0.5557,
+      "step": 3913
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.841302491480118,
+      "learning_rate": 8.790667570066941e-06,
+      "loss": 0.6534,
+      "step": 3914
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6025201812345489,
+      "learning_rate": 8.78999479268432e-06,
+      "loss": 0.4833,
+      "step": 3915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.814955347674252,
+      "learning_rate": 8.789321853973733e-06,
+      "loss": 0.677,
+      "step": 3916
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7262445888644592,
+      "learning_rate": 8.788648753963822e-06,
+      "loss": 0.6205,
+      "step": 3917
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7006669001515851,
+      "learning_rate": 8.78797549268324e-06,
+      "loss": 0.5315,
+      "step": 3918
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8567688768015841,
+      "learning_rate": 8.787302070160649e-06,
+      "loss": 0.6288,
+      "step": 3919
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5519427008134669,
+      "learning_rate": 8.786628486424707e-06,
+      "loss": 0.5202,
+      "step": 3920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7393250649816075,
+      "learning_rate": 8.785954741504093e-06,
+      "loss": 0.6294,
+      "step": 3921
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6297069959304996,
+      "learning_rate": 8.785280835427484e-06,
+      "loss": 0.4981,
+      "step": 3922
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6528378035469024,
+      "learning_rate": 8.784606768223566e-06,
+      "loss": 0.5373,
+      "step": 3923
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6227212886003048,
+      "learning_rate": 8.783932539921032e-06,
+      "loss": 0.5402,
+      "step": 3924
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9937598494539905,
+      "learning_rate": 8.78325815054858e-06,
+      "loss": 0.599,
+      "step": 3925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6215995681617628,
+      "learning_rate": 8.782583600134918e-06,
+      "loss": 0.5335,
+      "step": 3926
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5380554764527031,
+      "learning_rate": 8.78190888870876e-06,
+      "loss": 0.4775,
+      "step": 3927
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7026124895736224,
+      "learning_rate": 8.781234016298828e-06,
+      "loss": 0.5746,
+      "step": 3928
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8598823617613709,
+      "learning_rate": 8.780558982933844e-06,
+      "loss": 0.6042,
+      "step": 3929
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8975364148059026,
+      "learning_rate": 8.779883788642547e-06,
+      "loss": 0.6304,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.541520899074446,
+      "learning_rate": 8.779208433453674e-06,
+      "loss": 0.5239,
+      "step": 3931
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.710982995919678,
+      "learning_rate": 8.778532917395976e-06,
+      "loss": 0.6096,
+      "step": 3932
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5236664120263389,
+      "learning_rate": 8.777857240498207e-06,
+      "loss": 0.4722,
+      "step": 3933
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7267115813270354,
+      "learning_rate": 8.777181402789126e-06,
+      "loss": 0.6277,
+      "step": 3934
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7040182474563481,
+      "learning_rate": 8.776505404297505e-06,
+      "loss": 0.6227,
+      "step": 3935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5660328735296014,
+      "learning_rate": 8.775829245052115e-06,
+      "loss": 0.516,
+      "step": 3936
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9056767207848451,
+      "learning_rate": 8.775152925081738e-06,
+      "loss": 0.6336,
+      "step": 3937
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.565275260062294,
+      "learning_rate": 8.774476444415168e-06,
+      "loss": 0.4978,
+      "step": 3938
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8544456932806831,
+      "learning_rate": 8.773799803081195e-06,
+      "loss": 0.6114,
+      "step": 3939
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7471981205121471,
+      "learning_rate": 8.773123001108626e-06,
+      "loss": 0.5687,
+      "step": 3940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7788288350058924,
+      "learning_rate": 8.772446038526267e-06,
+      "loss": 0.6382,
+      "step": 3941
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7295741329091895,
+      "learning_rate": 8.771768915362935e-06,
+      "loss": 0.5073,
+      "step": 3942
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7935413544226886,
+      "learning_rate": 8.771091631647453e-06,
+      "loss": 0.5818,
+      "step": 3943
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9755369019121798,
+      "learning_rate": 8.770414187408652e-06,
+      "loss": 0.6647,
+      "step": 3944
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8530256658483412,
+      "learning_rate": 8.769736582675366e-06,
+      "loss": 0.6504,
+      "step": 3945
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7088817180572102,
+      "learning_rate": 8.76905881747644e-06,
+      "loss": 0.6092,
+      "step": 3946
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6121614392287799,
+      "learning_rate": 8.768380891840725e-06,
+      "loss": 0.5421,
+      "step": 3947
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8196487573531597,
+      "learning_rate": 8.767702805797077e-06,
+      "loss": 0.6215,
+      "step": 3948
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6000699850241337,
+      "learning_rate": 8.767024559374361e-06,
+      "loss": 0.5263,
+      "step": 3949
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6605492076345316,
+      "learning_rate": 8.766346152601448e-06,
+      "loss": 0.53,
+      "step": 3950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6352723081872792,
+      "learning_rate": 8.765667585507213e-06,
+      "loss": 0.5504,
+      "step": 3951
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7852571521634043,
+      "learning_rate": 8.764988858120543e-06,
+      "loss": 0.641,
+      "step": 3952
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7361326200022199,
+      "learning_rate": 8.764309970470328e-06,
+      "loss": 0.6594,
+      "step": 3953
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6980304144144914,
+      "learning_rate": 8.763630922585466e-06,
+      "loss": 0.6162,
+      "step": 3954
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6626658153353427,
+      "learning_rate": 8.762951714494864e-06,
+      "loss": 0.5584,
+      "step": 3955
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6939803321819855,
+      "learning_rate": 8.762272346227434e-06,
+      "loss": 0.5594,
+      "step": 3956
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7735382705157501,
+      "learning_rate": 8.761592817812088e-06,
+      "loss": 0.6308,
+      "step": 3957
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8090253150963491,
+      "learning_rate": 8.76091312927776e-06,
+      "loss": 0.6188,
+      "step": 3958
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6500013847522571,
+      "learning_rate": 8.760233280653376e-06,
+      "loss": 0.5905,
+      "step": 3959
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6895783995054142,
+      "learning_rate": 8.75955327196788e-06,
+      "loss": 0.5169,
+      "step": 3960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5468811719940511,
+      "learning_rate": 8.758873103250212e-06,
+      "loss": 0.4966,
+      "step": 3961
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7887595464839658,
+      "learning_rate": 8.758192774529328e-06,
+      "loss": 0.6637,
+      "step": 3962
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7138163027372428,
+      "learning_rate": 8.757512285834189e-06,
+      "loss": 0.6375,
+      "step": 3963
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5599752231453183,
+      "learning_rate": 8.756831637193757e-06,
+      "loss": 0.5507,
+      "step": 3964
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6959909932970919,
+      "learning_rate": 8.756150828637009e-06,
+      "loss": 0.5288,
+      "step": 3965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7716403889418847,
+      "learning_rate": 8.755469860192922e-06,
+      "loss": 0.5618,
+      "step": 3966
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7329369573913461,
+      "learning_rate": 8.754788731890485e-06,
+      "loss": 0.5584,
+      "step": 3967
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7316888058876729,
+      "learning_rate": 8.75410744375869e-06,
+      "loss": 0.6067,
+      "step": 3968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5722472749697322,
+      "learning_rate": 8.753425995826536e-06,
+      "loss": 0.5167,
+      "step": 3969
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5915106949795744,
+      "learning_rate": 8.752744388123033e-06,
+      "loss": 0.5552,
+      "step": 3970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8081715364208715,
+      "learning_rate": 8.752062620677193e-06,
+      "loss": 0.6373,
+      "step": 3971
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5530192994616905,
+      "learning_rate": 8.751380693518038e-06,
+      "loss": 0.4732,
+      "step": 3972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5950488609759187,
+      "learning_rate": 8.750698606674594e-06,
+      "loss": 0.5268,
+      "step": 3973
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5749555327536146,
+      "learning_rate": 8.750016360175895e-06,
+      "loss": 0.5284,
+      "step": 3974
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6571820300636001,
+      "learning_rate": 8.749333954050984e-06,
+      "loss": 0.6087,
+      "step": 3975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6477038814961695,
+      "learning_rate": 8.748651388328906e-06,
+      "loss": 0.535,
+      "step": 3976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6708654495194948,
+      "learning_rate": 8.74796866303872e-06,
+      "loss": 0.568,
+      "step": 3977
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7169526277350821,
+      "learning_rate": 8.747285778209483e-06,
+      "loss": 0.5198,
+      "step": 3978
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7714452218504663,
+      "learning_rate": 8.746602733870266e-06,
+      "loss": 0.6246,
+      "step": 3979
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7184550479829447,
+      "learning_rate": 8.745919530050143e-06,
+      "loss": 0.5991,
+      "step": 3980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8511114247254772,
+      "learning_rate": 8.745236166778194e-06,
+      "loss": 0.6768,
+      "step": 3981
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5887325762540285,
+      "learning_rate": 8.744552644083512e-06,
+      "loss": 0.512,
+      "step": 3982
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7314092283086151,
+      "learning_rate": 8.743868961995188e-06,
+      "loss": 0.4886,
+      "step": 3983
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6158300606820595,
+      "learning_rate": 8.743185120542326e-06,
+      "loss": 0.5424,
+      "step": 3984
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6551168389002009,
+      "learning_rate": 8.742501119754037e-06,
+      "loss": 0.5373,
+      "step": 3985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.573471529393463,
+      "learning_rate": 8.741816959659432e-06,
+      "loss": 0.5122,
+      "step": 3986
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6910089267339331,
+      "learning_rate": 8.741132640287639e-06,
+      "loss": 0.5785,
+      "step": 3987
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6899036001471719,
+      "learning_rate": 8.740448161667783e-06,
+      "loss": 0.5408,
+      "step": 3988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7321408321882751,
+      "learning_rate": 8.739763523829001e-06,
+      "loss": 0.5501,
+      "step": 3989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7112818807731353,
+      "learning_rate": 8.739078726800436e-06,
+      "loss": 0.5512,
+      "step": 3990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1155330792726104,
+      "learning_rate": 8.73839377061124e-06,
+      "loss": 0.6421,
+      "step": 3991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5606726973898964,
+      "learning_rate": 8.737708655290564e-06,
+      "loss": 0.5187,
+      "step": 3992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7522452464776879,
+      "learning_rate": 8.737023380867578e-06,
+      "loss": 0.5994,
+      "step": 3993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9150160363986606,
+      "learning_rate": 8.736337947371448e-06,
+      "loss": 0.6838,
+      "step": 3994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7507945652532474,
+      "learning_rate": 8.73565235483135e-06,
+      "loss": 0.5765,
+      "step": 3995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5785208228204933,
+      "learning_rate": 8.73496660327647e-06,
+      "loss": 0.5373,
+      "step": 3996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5357985068475164,
+      "learning_rate": 8.734280692735995e-06,
+      "loss": 0.5086,
+      "step": 3997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.572289222696746,
+      "learning_rate": 8.733594623239125e-06,
+      "loss": 0.5474,
+      "step": 3998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7119114566450648,
+      "learning_rate": 8.732908394815063e-06,
+      "loss": 0.6178,
+      "step": 3999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7543084118528486,
+      "learning_rate": 8.732222007493019e-06,
+      "loss": 0.5496,
+      "step": 4000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6542244018315772,
+      "learning_rate": 8.73153546130221e-06,
+      "loss": 0.5878,
+      "step": 4001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7133576120049314,
+      "learning_rate": 8.730848756271862e-06,
+      "loss": 0.639,
+      "step": 4002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7365670830170392,
+      "learning_rate": 8.730161892431204e-06,
+      "loss": 0.5871,
+      "step": 4003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8981157433936465,
+      "learning_rate": 8.729474869809474e-06,
+      "loss": 0.6407,
+      "step": 4004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5689923401973566,
+      "learning_rate": 8.728787688435916e-06,
+      "loss": 0.5288,
+      "step": 4005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6219527165380645,
+      "learning_rate": 8.728100348339783e-06,
+      "loss": 0.5202,
+      "step": 4006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.630844260409958,
+      "learning_rate": 8.72741284955033e-06,
+      "loss": 0.5255,
+      "step": 4007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7227527979511357,
+      "learning_rate": 8.726725192096824e-06,
+      "loss": 0.5761,
+      "step": 4008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6202240367694997,
+      "learning_rate": 8.726037376008536e-06,
+      "loss": 0.4882,
+      "step": 4009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6559508357445297,
+      "learning_rate": 8.72534940131474e-06,
+      "loss": 0.5337,
+      "step": 4010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6405642393842376,
+      "learning_rate": 8.724661268044728e-06,
+      "loss": 0.48,
+      "step": 4011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6600537617560803,
+      "learning_rate": 8.723972976227788e-06,
+      "loss": 0.5125,
+      "step": 4012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6226580315348801,
+      "learning_rate": 8.723284525893219e-06,
+      "loss": 0.5397,
+      "step": 4013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6788184469513424,
+      "learning_rate": 8.722595917070324e-06,
+      "loss": 0.5355,
+      "step": 4014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5946836383408967,
+      "learning_rate": 8.721907149788416e-06,
+      "loss": 0.5817,
+      "step": 4015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6621622427559274,
+      "learning_rate": 8.721218224076815e-06,
+      "loss": 0.5309,
+      "step": 4016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6623768717483649,
+      "learning_rate": 8.720529139964846e-06,
+      "loss": 0.5924,
+      "step": 4017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8514805777793002,
+      "learning_rate": 8.71983989748184e-06,
+      "loss": 0.62,
+      "step": 4018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.744568234209917,
+      "learning_rate": 8.719150496657138e-06,
+      "loss": 0.6015,
+      "step": 4019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8172516781128346,
+      "learning_rate": 8.718460937520083e-06,
+      "loss": 0.6092,
+      "step": 4020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7186431297226278,
+      "learning_rate": 8.717771220100028e-06,
+      "loss": 0.6158,
+      "step": 4021
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5596547073665259,
+      "learning_rate": 8.717081344426332e-06,
+      "loss": 0.5079,
+      "step": 4022
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7277117309897767,
+      "learning_rate": 8.716391310528362e-06,
+      "loss": 0.6234,
+      "step": 4023
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5581829383311955,
+      "learning_rate": 8.71570111843549e-06,
+      "loss": 0.532,
+      "step": 4024
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8546352303153767,
+      "learning_rate": 8.715010768177096e-06,
+      "loss": 0.6008,
+      "step": 4025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6431617435741185,
+      "learning_rate": 8.714320259782564e-06,
+      "loss": 0.5458,
+      "step": 4026
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.871148495129938,
+      "learning_rate": 8.713629593281287e-06,
+      "loss": 0.6525,
+      "step": 4027
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5651827233493597,
+      "learning_rate": 8.712938768702667e-06,
+      "loss": 0.5311,
+      "step": 4028
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6910219990548543,
+      "learning_rate": 8.712247786076107e-06,
+      "loss": 0.5667,
+      "step": 4029
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6234576595087924,
+      "learning_rate": 8.711556645431023e-06,
+      "loss": 0.5451,
+      "step": 4030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7584569646657977,
+      "learning_rate": 8.710865346796831e-06,
+      "loss": 0.6466,
+      "step": 4031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7033104790397724,
+      "learning_rate": 8.71017389020296e-06,
+      "loss": 0.6485,
+      "step": 4032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6413070074062487,
+      "learning_rate": 8.709482275678844e-06,
+      "loss": 0.5302,
+      "step": 4033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9181804849062695,
+      "learning_rate": 8.70879050325392e-06,
+      "loss": 0.6107,
+      "step": 4034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.614806654088529,
+      "learning_rate": 8.708098572957634e-06,
+      "loss": 0.5341,
+      "step": 4035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6211892196371582,
+      "learning_rate": 8.707406484819445e-06,
+      "loss": 0.5617,
+      "step": 4036
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7178186116677533,
+      "learning_rate": 8.706714238868808e-06,
+      "loss": 0.5269,
+      "step": 4037
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.667514713153214,
+      "learning_rate": 8.70602183513519e-06,
+      "loss": 0.54,
+      "step": 4038
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7723771910555213,
+      "learning_rate": 8.705329273648065e-06,
+      "loss": 0.6125,
+      "step": 4039
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7126742556076741,
+      "learning_rate": 8.704636554436913e-06,
+      "loss": 0.5893,
+      "step": 4040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8399445894924042,
+      "learning_rate": 8.703943677531223e-06,
+      "loss": 0.5503,
+      "step": 4041
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8695449332268784,
+      "learning_rate": 8.703250642960486e-06,
+      "loss": 0.6168,
+      "step": 4042
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7359725643449906,
+      "learning_rate": 8.702557450754201e-06,
+      "loss": 0.607,
+      "step": 4043
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7472907414402942,
+      "learning_rate": 8.701864100941879e-06,
+      "loss": 0.5726,
+      "step": 4044
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6833697075096975,
+      "learning_rate": 8.701170593553032e-06,
+      "loss": 0.5109,
+      "step": 4045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8930552361383712,
+      "learning_rate": 8.70047692861718e-06,
+      "loss": 0.6676,
+      "step": 4046
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6422085922646437,
+      "learning_rate": 8.69978310616385e-06,
+      "loss": 0.5863,
+      "step": 4047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.656960316599447,
+      "learning_rate": 8.699089126222576e-06,
+      "loss": 0.5375,
+      "step": 4048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.676279498152905,
+      "learning_rate": 8.698394988822898e-06,
+      "loss": 0.5397,
+      "step": 4049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6276087711034353,
+      "learning_rate": 8.697700693994363e-06,
+      "loss": 0.518,
+      "step": 4050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6371970251125455,
+      "learning_rate": 8.697006241766527e-06,
+      "loss": 0.5037,
+      "step": 4051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6893087387319538,
+      "learning_rate": 8.696311632168949e-06,
+      "loss": 0.5757,
+      "step": 4052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6162191817085644,
+      "learning_rate": 8.695616865231195e-06,
+      "loss": 0.5383,
+      "step": 4053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7829475087524488,
+      "learning_rate": 8.694921940982843e-06,
+      "loss": 0.5308,
+      "step": 4054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8307525867404142,
+      "learning_rate": 8.694226859453469e-06,
+      "loss": 0.616,
+      "step": 4055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9620411113643442,
+      "learning_rate": 8.693531620672664e-06,
+      "loss": 0.6843,
+      "step": 4056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6578430396849646,
+      "learning_rate": 8.69283622467002e-06,
+      "loss": 0.5298,
+      "step": 4057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5854872539494609,
+      "learning_rate": 8.69214067147514e-06,
+      "loss": 0.5428,
+      "step": 4058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6192794312437374,
+      "learning_rate": 8.69144496111763e-06,
+      "loss": 0.5471,
+      "step": 4059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.61017652142063,
+      "learning_rate": 8.690749093627102e-06,
+      "loss": 0.5551,
+      "step": 4060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7824772911012285,
+      "learning_rate": 8.690053069033182e-06,
+      "loss": 0.5582,
+      "step": 4061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6472949686667087,
+      "learning_rate": 8.689356887365492e-06,
+      "loss": 0.5841,
+      "step": 4062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7541421204474587,
+      "learning_rate": 8.68866054865367e-06,
+      "loss": 0.6478,
+      "step": 4063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5682369069903911,
+      "learning_rate": 8.687964052927354e-06,
+      "loss": 0.5286,
+      "step": 4064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.843779305332011,
+      "learning_rate": 8.687267400216195e-06,
+      "loss": 0.6123,
+      "step": 4065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8020967312698006,
+      "learning_rate": 8.686570590549845e-06,
+      "loss": 0.6425,
+      "step": 4066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7308260065605556,
+      "learning_rate": 8.685873623957966e-06,
+      "loss": 0.6648,
+      "step": 4067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5821694481684088,
+      "learning_rate": 8.685176500470225e-06,
+      "loss": 0.5071,
+      "step": 4068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6087344874233439,
+      "learning_rate": 8.684479220116296e-06,
+      "loss": 0.5287,
+      "step": 4069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6595120773884432,
+      "learning_rate": 8.683781782925861e-06,
+      "loss": 0.5019,
+      "step": 4070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3103340262646546,
+      "learning_rate": 8.683084188928608e-06,
+      "loss": 0.6074,
+      "step": 4071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7099659840423111,
+      "learning_rate": 8.68238643815423e-06,
+      "loss": 0.6247,
+      "step": 4072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7628189058571455,
+      "learning_rate": 8.681688530632429e-06,
+      "loss": 0.6058,
+      "step": 4073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6555232246497514,
+      "learning_rate": 8.680990466392912e-06,
+      "loss": 0.5221,
+      "step": 4074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3898887696892253,
+      "learning_rate": 8.680292245465392e-06,
+      "loss": 0.5638,
+      "step": 4075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6140742381851138,
+      "learning_rate": 8.679593867879592e-06,
+      "loss": 0.5619,
+      "step": 4076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7637791651443466,
+      "learning_rate": 8.67889533366524e-06,
+      "loss": 0.6555,
+      "step": 4077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7524925106588084,
+      "learning_rate": 8.67819664285207e-06,
+      "loss": 0.6099,
+      "step": 4078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6666990954598578,
+      "learning_rate": 8.677497795469823e-06,
+      "loss": 0.603,
+      "step": 4079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7646750122609234,
+      "learning_rate": 8.676798791548246e-06,
+      "loss": 0.612,
+      "step": 4080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5656202450360042,
+      "learning_rate": 8.676099631117094e-06,
+      "loss": 0.5268,
+      "step": 4081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6034868943379306,
+      "learning_rate": 8.675400314206126e-06,
+      "loss": 0.5224,
+      "step": 4082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8118847928983691,
+      "learning_rate": 8.674700840845114e-06,
+      "loss": 0.6165,
+      "step": 4083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.641151788610495,
+      "learning_rate": 8.674001211063829e-06,
+      "loss": 0.5163,
+      "step": 4084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6908073121201854,
+      "learning_rate": 8.673301424892053e-06,
+      "loss": 0.5475,
+      "step": 4085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6066042040860486,
+      "learning_rate": 8.672601482359575e-06,
+      "loss": 0.5278,
+      "step": 4086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6191794521350743,
+      "learning_rate": 8.671901383496186e-06,
+      "loss": 0.5292,
+      "step": 4087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7959269387054396,
+      "learning_rate": 8.671201128331687e-06,
+      "loss": 0.6259,
+      "step": 4088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8350898812561149,
+      "learning_rate": 8.67050071689589e-06,
+      "loss": 0.6722,
+      "step": 4089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7523124281825234,
+      "learning_rate": 8.669800149218607e-06,
+      "loss": 0.584,
+      "step": 4090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6330707856985351,
+      "learning_rate": 8.66909942532966e-06,
+      "loss": 0.5714,
+      "step": 4091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5500080185137677,
+      "learning_rate": 8.668398545258872e-06,
+      "loss": 0.4706,
+      "step": 4092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7440400781920494,
+      "learning_rate": 8.667697509036081e-06,
+      "loss": 0.5917,
+      "step": 4093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7571174341942223,
+      "learning_rate": 8.666996316691129e-06,
+      "loss": 0.594,
+      "step": 4094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5886212622941261,
+      "learning_rate": 8.66629496825386e-06,
+      "loss": 0.5436,
+      "step": 4095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8535177447626139,
+      "learning_rate": 8.665593463754131e-06,
+      "loss": 0.6544,
+      "step": 4096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6216790588125976,
+      "learning_rate": 8.664891803221802e-06,
+      "loss": 0.5401,
+      "step": 4097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6166756233687279,
+      "learning_rate": 8.66418998668674e-06,
+      "loss": 0.5657,
+      "step": 4098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6877739594834584,
+      "learning_rate": 8.66348801417882e-06,
+      "loss": 0.5408,
+      "step": 4099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6640935474929383,
+      "learning_rate": 8.66278588572792e-06,
+      "loss": 0.5717,
+      "step": 4100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7394274919281051,
+      "learning_rate": 8.66208360136393e-06,
+      "loss": 0.5924,
+      "step": 4101
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7655150255633112,
+      "learning_rate": 8.661381161116745e-06,
+      "loss": 0.5503,
+      "step": 4102
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6972852218878891,
+      "learning_rate": 8.660678565016263e-06,
+      "loss": 0.5819,
+      "step": 4103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7388092332104087,
+      "learning_rate": 8.659975813092393e-06,
+      "loss": 0.5306,
+      "step": 4104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7989478190168652,
+      "learning_rate": 8.659272905375047e-06,
+      "loss": 0.588,
+      "step": 4105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6038444889572419,
+      "learning_rate": 8.658569841894148e-06,
+      "loss": 0.5637,
+      "step": 4106
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7578801116926419,
+      "learning_rate": 8.657866622679623e-06,
+      "loss": 0.5777,
+      "step": 4107
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6616704042341236,
+      "learning_rate": 8.657163247761402e-06,
+      "loss": 0.5615,
+      "step": 4108
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5650059545608774,
+      "learning_rate": 8.656459717169429e-06,
+      "loss": 0.4692,
+      "step": 4109
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6599071762573367,
+      "learning_rate": 8.65575603093365e-06,
+      "loss": 0.4731,
+      "step": 4110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9157734941245574,
+      "learning_rate": 8.655052189084022e-06,
+      "loss": 0.6301,
+      "step": 4111
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.806251991510044,
+      "learning_rate": 8.654348191650499e-06,
+      "loss": 0.5543,
+      "step": 4112
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4774434268320176,
+      "learning_rate": 8.653644038663051e-06,
+      "loss": 0.6115,
+      "step": 4113
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5390721471166144,
+      "learning_rate": 8.652939730151653e-06,
+      "loss": 0.5035,
+      "step": 4114
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8317143946613599,
+      "learning_rate": 8.652235266146284e-06,
+      "loss": 0.6595,
+      "step": 4115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7913826290662731,
+      "learning_rate": 8.651530646676928e-06,
+      "loss": 0.6217,
+      "step": 4116
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.654443711056819,
+      "learning_rate": 8.650825871773586e-06,
+      "loss": 0.5208,
+      "step": 4117
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7653087047593622,
+      "learning_rate": 8.650120941466248e-06,
+      "loss": 0.6432,
+      "step": 4118
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.57040990562097,
+      "learning_rate": 8.649415855784929e-06,
+      "loss": 0.5459,
+      "step": 4119
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6527354173215035,
+      "learning_rate": 8.648710614759637e-06,
+      "loss": 0.5758,
+      "step": 4120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8138528256104013,
+      "learning_rate": 8.648005218420393e-06,
+      "loss": 0.588,
+      "step": 4121
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.613926470552269,
+      "learning_rate": 8.647299666797226e-06,
+      "loss": 0.5711,
+      "step": 4122
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7165810152569417,
+      "learning_rate": 8.646593959920166e-06,
+      "loss": 0.6382,
+      "step": 4123
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8305220534314462,
+      "learning_rate": 8.645888097819255e-06,
+      "loss": 0.6302,
+      "step": 4124
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.610496215479003,
+      "learning_rate": 8.64518208052454e-06,
+      "loss": 0.5477,
+      "step": 4125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5918479460318838,
+      "learning_rate": 8.644475908066067e-06,
+      "loss": 0.5531,
+      "step": 4126
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8880709961866834,
+      "learning_rate": 8.643769580473905e-06,
+      "loss": 0.6025,
+      "step": 4127
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7413602925646209,
+      "learning_rate": 8.643063097778115e-06,
+      "loss": 0.6323,
+      "step": 4128
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9621443936566458,
+      "learning_rate": 8.642356460008769e-06,
+      "loss": 0.6034,
+      "step": 4129
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5911236851883683,
+      "learning_rate": 8.641649667195951e-06,
+      "loss": 0.5578,
+      "step": 4130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5623687100068212,
+      "learning_rate": 8.640942719369741e-06,
+      "loss": 0.5149,
+      "step": 4131
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.695412717064727,
+      "learning_rate": 8.640235616560234e-06,
+      "loss": 0.6133,
+      "step": 4132
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9097797461029631,
+      "learning_rate": 8.639528358797531e-06,
+      "loss": 0.5955,
+      "step": 4133
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7032956044712427,
+      "learning_rate": 8.638820946111735e-06,
+      "loss": 0.5135,
+      "step": 4134
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6102219813605976,
+      "learning_rate": 8.638113378532958e-06,
+      "loss": 0.5143,
+      "step": 4135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8384911835924354,
+      "learning_rate": 8.637405656091323e-06,
+      "loss": 0.5976,
+      "step": 4136
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9153487422463845,
+      "learning_rate": 8.63669777881695e-06,
+      "loss": 0.6568,
+      "step": 4137
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5797435423856463,
+      "learning_rate": 8.635989746739977e-06,
+      "loss": 0.509,
+      "step": 4138
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6914392041462173,
+      "learning_rate": 8.635281559890536e-06,
+      "loss": 0.5441,
+      "step": 4139
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7410179945760179,
+      "learning_rate": 8.634573218298778e-06,
+      "loss": 0.5199,
+      "step": 4140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6525900131655286,
+      "learning_rate": 8.633864721994852e-06,
+      "loss": 0.5494,
+      "step": 4141
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5900897350075014,
+      "learning_rate": 8.633156071008917e-06,
+      "loss": 0.5399,
+      "step": 4142
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8346246161729853,
+      "learning_rate": 8.632447265371137e-06,
+      "loss": 0.595,
+      "step": 4143
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9441855549004021,
+      "learning_rate": 8.631738305111686e-06,
+      "loss": 0.6699,
+      "step": 4144
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6496769937815052,
+      "learning_rate": 8.63102919026074e-06,
+      "loss": 0.5157,
+      "step": 4145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6931401396567558,
+      "learning_rate": 8.630319920848485e-06,
+      "loss": 0.5379,
+      "step": 4146
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.650106143481835,
+      "learning_rate": 8.62961049690511e-06,
+      "loss": 0.587,
+      "step": 4147
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5441884620634613,
+      "learning_rate": 8.628900918460816e-06,
+      "loss": 0.5076,
+      "step": 4148
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.791754450685671,
+      "learning_rate": 8.628191185545806e-06,
+      "loss": 0.5994,
+      "step": 4149
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7597955361054618,
+      "learning_rate": 8.627481298190292e-06,
+      "loss": 0.6826,
+      "step": 4150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8251425342384988,
+      "learning_rate": 8.62677125642449e-06,
+      "loss": 0.6157,
+      "step": 4151
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.650528724218808,
+      "learning_rate": 8.626061060278627e-06,
+      "loss": 0.555,
+      "step": 4152
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7189530729765402,
+      "learning_rate": 8.62535070978293e-06,
+      "loss": 0.5655,
+      "step": 4153
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7606499760756236,
+      "learning_rate": 8.62464020496764e-06,
+      "loss": 0.6268,
+      "step": 4154
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5732538416130583,
+      "learning_rate": 8.623929545862997e-06,
+      "loss": 0.4855,
+      "step": 4155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7301025580970828,
+      "learning_rate": 8.623218732499254e-06,
+      "loss": 0.6045,
+      "step": 4156
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7152390100076712,
+      "learning_rate": 8.622507764906669e-06,
+      "loss": 0.5478,
+      "step": 4157
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7711394971385065,
+      "learning_rate": 8.621796643115503e-06,
+      "loss": 0.6485,
+      "step": 4158
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.693863683938519,
+      "learning_rate": 8.621085367156028e-06,
+      "loss": 0.4978,
+      "step": 4159
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7262533239298832,
+      "learning_rate": 8.62037393705852e-06,
+      "loss": 0.6006,
+      "step": 4160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7916697470361381,
+      "learning_rate": 8.619662352853264e-06,
+      "loss": 0.6171,
+      "step": 4161
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6121833954315268,
+      "learning_rate": 8.618950614570547e-06,
+      "loss": 0.5233,
+      "step": 4162
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5597290368600105,
+      "learning_rate": 8.618238722240668e-06,
+      "loss": 0.4886,
+      "step": 4163
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7417499369103104,
+      "learning_rate": 8.617526675893928e-06,
+      "loss": 0.6362,
+      "step": 4164
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9988250437935274,
+      "learning_rate": 8.616814475560636e-06,
+      "loss": 0.6653,
+      "step": 4165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5158037139850548,
+      "learning_rate": 8.61610212127111e-06,
+      "loss": 0.4958,
+      "step": 4166
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7505200182807563,
+      "learning_rate": 8.615389613055673e-06,
+      "loss": 0.6192,
+      "step": 4167
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6833847468301292,
+      "learning_rate": 8.614676950944654e-06,
+      "loss": 0.6084,
+      "step": 4168
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7950884612137535,
+      "learning_rate": 8.613964134968387e-06,
+      "loss": 0.6446,
+      "step": 4169
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.760316459249358,
+      "learning_rate": 8.613251165157217e-06,
+      "loss": 0.6126,
+      "step": 4170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7438221263805592,
+      "learning_rate": 8.612538041541489e-06,
+      "loss": 0.6419,
+      "step": 4171
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6463356886247231,
+      "learning_rate": 8.611824764151561e-06,
+      "loss": 0.5119,
+      "step": 4172
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6826826851340493,
+      "learning_rate": 8.611111333017796e-06,
+      "loss": 0.6654,
+      "step": 4173
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7182546328168065,
+      "learning_rate": 8.61039774817056e-06,
+      "loss": 0.5875,
+      "step": 4174
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6464219189713521,
+      "learning_rate": 8.609684009640229e-06,
+      "loss": 0.5131,
+      "step": 4175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8153351876063932,
+      "learning_rate": 8.608970117457184e-06,
+      "loss": 0.6086,
+      "step": 4176
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7910516754668436,
+      "learning_rate": 8.608256071651815e-06,
+      "loss": 0.6165,
+      "step": 4177
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.656802740069538,
+      "learning_rate": 8.607541872254514e-06,
+      "loss": 0.5108,
+      "step": 4178
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6371077279264837,
+      "learning_rate": 8.606827519295684e-06,
+      "loss": 0.5285,
+      "step": 4179
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5539188929778869,
+      "learning_rate": 8.606113012805733e-06,
+      "loss": 0.5101,
+      "step": 4180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6357198748650659,
+      "learning_rate": 8.605398352815074e-06,
+      "loss": 0.5168,
+      "step": 4181
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7508241451132187,
+      "learning_rate": 8.604683539354129e-06,
+      "loss": 0.5726,
+      "step": 4182
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6215320432201875,
+      "learning_rate": 8.603968572453324e-06,
+      "loss": 0.5758,
+      "step": 4183
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8068590590914627,
+      "learning_rate": 8.603253452143093e-06,
+      "loss": 0.6111,
+      "step": 4184
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7141856617069584,
+      "learning_rate": 8.602538178453877e-06,
+      "loss": 0.5656,
+      "step": 4185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5908150107095695,
+      "learning_rate": 8.601822751416124e-06,
+      "loss": 0.5656,
+      "step": 4186
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7204489416727751,
+      "learning_rate": 8.601107171060286e-06,
+      "loss": 0.5973,
+      "step": 4187
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5675320397907174,
+      "learning_rate": 8.600391437416822e-06,
+      "loss": 0.5187,
+      "step": 4188
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6409800876569068,
+      "learning_rate": 8.599675550516201e-06,
+      "loss": 0.5155,
+      "step": 4189
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5860737309056203,
+      "learning_rate": 8.598959510388894e-06,
+      "loss": 0.5359,
+      "step": 4190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9123702174969488,
+      "learning_rate": 8.598243317065382e-06,
+      "loss": 0.6837,
+      "step": 4191
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8262264583018809,
+      "learning_rate": 8.597526970576148e-06,
+      "loss": 0.6649,
+      "step": 4192
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6748319481331477,
+      "learning_rate": 8.59681047095169e-06,
+      "loss": 0.595,
+      "step": 4193
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.676612574805558,
+      "learning_rate": 8.596093818222504e-06,
+      "loss": 0.5552,
+      "step": 4194
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.823835763345746,
+      "learning_rate": 8.595377012419093e-06,
+      "loss": 0.6093,
+      "step": 4195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8484589908110719,
+      "learning_rate": 8.594660053571972e-06,
+      "loss": 0.6066,
+      "step": 4196
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7217782534743693,
+      "learning_rate": 8.593942941711662e-06,
+      "loss": 0.6151,
+      "step": 4197
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5632304908788441,
+      "learning_rate": 8.593225676868685e-06,
+      "loss": 0.5254,
+      "step": 4198
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6662700969602309,
+      "learning_rate": 8.592508259073571e-06,
+      "loss": 0.5316,
+      "step": 4199
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.737610740156497,
+      "learning_rate": 8.591790688356863e-06,
+      "loss": 0.5141,
+      "step": 4200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7547099648272309,
+      "learning_rate": 8.5910729647491e-06,
+      "loss": 0.5791,
+      "step": 4201
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6084062881886371,
+      "learning_rate": 8.59035508828084e-06,
+      "loss": 0.5719,
+      "step": 4202
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5751780275211879,
+      "learning_rate": 8.589637058982634e-06,
+      "loss": 0.5317,
+      "step": 4203
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5713846862244606,
+      "learning_rate": 8.58891887688505e-06,
+      "loss": 0.5295,
+      "step": 4204
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8001175862986285,
+      "learning_rate": 8.588200542018658e-06,
+      "loss": 0.5774,
+      "step": 4205
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6770304915731522,
+      "learning_rate": 8.587482054414036e-06,
+      "loss": 0.5484,
+      "step": 4206
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7087752401044223,
+      "learning_rate": 8.586763414101766e-06,
+      "loss": 0.5347,
+      "step": 4207
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7071601500404803,
+      "learning_rate": 8.586044621112442e-06,
+      "loss": 0.5589,
+      "step": 4208
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8252986939014532,
+      "learning_rate": 8.585325675476655e-06,
+      "loss": 0.5873,
+      "step": 4209
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6291140667432583,
+      "learning_rate": 8.58460657722501e-06,
+      "loss": 0.5605,
+      "step": 4210
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5163097025465593,
+      "learning_rate": 8.58388732638812e-06,
+      "loss": 0.5118,
+      "step": 4211
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8195100101489301,
+      "learning_rate": 8.583167922996598e-06,
+      "loss": 0.5969,
+      "step": 4212
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6322556113774183,
+      "learning_rate": 8.58244836708107e-06,
+      "loss": 0.617,
+      "step": 4213
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.587528861532704,
+      "learning_rate": 8.581728658672159e-06,
+      "loss": 0.4921,
+      "step": 4214
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5784900038882096,
+      "learning_rate": 8.581008797800506e-06,
+      "loss": 0.524,
+      "step": 4215
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6700150476268066,
+      "learning_rate": 8.580288784496752e-06,
+      "loss": 0.5758,
+      "step": 4216
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5539846050468967,
+      "learning_rate": 8.579568618791547e-06,
+      "loss": 0.4944,
+      "step": 4217
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7533418118582415,
+      "learning_rate": 8.578848300715542e-06,
+      "loss": 0.5857,
+      "step": 4218
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5890395111804739,
+      "learning_rate": 8.578127830299404e-06,
+      "loss": 0.5226,
+      "step": 4219
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6189759451509745,
+      "learning_rate": 8.577407207573795e-06,
+      "loss": 0.5422,
+      "step": 4220
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7313971748589089,
+      "learning_rate": 8.576686432569394e-06,
+      "loss": 0.5291,
+      "step": 4221
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6457195678984613,
+      "learning_rate": 8.575965505316883e-06,
+      "loss": 0.6376,
+      "step": 4222
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6017810013002097,
+      "learning_rate": 8.575244425846945e-06,
+      "loss": 0.522,
+      "step": 4223
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.665085071496339,
+      "learning_rate": 8.574523194190279e-06,
+      "loss": 0.5689,
+      "step": 4224
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5874504662682642,
+      "learning_rate": 8.573801810377581e-06,
+      "loss": 0.5589,
+      "step": 4225
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7687996514174105,
+      "learning_rate": 8.573080274439562e-06,
+      "loss": 0.5539,
+      "step": 4226
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.11924188433432,
+      "learning_rate": 8.572358586406933e-06,
+      "loss": 0.6066,
+      "step": 4227
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6729692267052756,
+      "learning_rate": 8.571636746310414e-06,
+      "loss": 0.4876,
+      "step": 4228
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.666474001354608,
+      "learning_rate": 8.570914754180733e-06,
+      "loss": 0.5208,
+      "step": 4229
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7473142683705336,
+      "learning_rate": 8.57019261004862e-06,
+      "loss": 0.5898,
+      "step": 4230
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6328917397470593,
+      "learning_rate": 8.56947031394482e-06,
+      "loss": 0.5285,
+      "step": 4231
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7986635809428788,
+      "learning_rate": 8.568747865900071e-06,
+      "loss": 0.5585,
+      "step": 4232
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7868504692724377,
+      "learning_rate": 8.56802526594513e-06,
+      "loss": 0.5947,
+      "step": 4233
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6468800138600855,
+      "learning_rate": 8.567302514110758e-06,
+      "loss": 0.5143,
+      "step": 4234
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7471634788203108,
+      "learning_rate": 8.566579610427715e-06,
+      "loss": 0.5681,
+      "step": 4235
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6310562666237889,
+      "learning_rate": 8.565856554926776e-06,
+      "loss": 0.5338,
+      "step": 4236
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6684379571259103,
+      "learning_rate": 8.565133347638719e-06,
+      "loss": 0.5026,
+      "step": 4237
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7049264857095631,
+      "learning_rate": 8.564409988594327e-06,
+      "loss": 0.5413,
+      "step": 4238
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6182126875094807,
+      "learning_rate": 8.563686477824392e-06,
+      "loss": 0.5407,
+      "step": 4239
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7842127618823762,
+      "learning_rate": 8.562962815359712e-06,
+      "loss": 0.6675,
+      "step": 4240
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7562142014068194,
+      "learning_rate": 8.56223900123109e-06,
+      "loss": 0.6016,
+      "step": 4241
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.880586079244348,
+      "learning_rate": 8.561515035469337e-06,
+      "loss": 0.6347,
+      "step": 4242
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7206524348861861,
+      "learning_rate": 8.56079091810527e-06,
+      "loss": 0.5485,
+      "step": 4243
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6043269227553943,
+      "learning_rate": 8.560066649169713e-06,
+      "loss": 0.546,
+      "step": 4244
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7613203813613829,
+      "learning_rate": 8.559342228693494e-06,
+      "loss": 0.6181,
+      "step": 4245
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7318250387236054,
+      "learning_rate": 8.558617656707452e-06,
+      "loss": 0.5911,
+      "step": 4246
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7224439786834228,
+      "learning_rate": 8.557892933242427e-06,
+      "loss": 0.6229,
+      "step": 4247
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7243203917415874,
+      "learning_rate": 8.55716805832927e-06,
+      "loss": 0.5974,
+      "step": 4248
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.626188156046898,
+      "learning_rate": 8.556443031998834e-06,
+      "loss": 0.528,
+      "step": 4249
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7931637331244187,
+      "learning_rate": 8.555717854281984e-06,
+      "loss": 0.6025,
+      "step": 4250
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6933561182954645,
+      "learning_rate": 8.554992525209589e-06,
+      "loss": 0.5412,
+      "step": 4251
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7755683091433357,
+      "learning_rate": 8.55426704481252e-06,
+      "loss": 0.6288,
+      "step": 4252
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.585608131688245,
+      "learning_rate": 8.55354141312166e-06,
+      "loss": 0.5274,
+      "step": 4253
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5471904085494139,
+      "learning_rate": 8.552815630167901e-06,
+      "loss": 0.5133,
+      "step": 4254
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6381377826857912,
+      "learning_rate": 8.552089695982132e-06,
+      "loss": 0.5194,
+      "step": 4255
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5628675477002059,
+      "learning_rate": 8.551363610595253e-06,
+      "loss": 0.4989,
+      "step": 4256
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6824626597554021,
+      "learning_rate": 8.550637374038178e-06,
+      "loss": 0.572,
+      "step": 4257
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.757329896163903,
+      "learning_rate": 8.549910986341813e-06,
+      "loss": 0.5717,
+      "step": 4258
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6578028395022709,
+      "learning_rate": 8.549184447537082e-06,
+      "loss": 0.576,
+      "step": 4259
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5724289711452756,
+      "learning_rate": 8.548457757654912e-06,
+      "loss": 0.515,
+      "step": 4260
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5910746454046204,
+      "learning_rate": 8.547730916726234e-06,
+      "loss": 0.5374,
+      "step": 4261
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.944792849147322,
+      "learning_rate": 8.547003924781988e-06,
+      "loss": 0.6537,
+      "step": 4262
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5978913596841029,
+      "learning_rate": 8.546276781853117e-06,
+      "loss": 0.5498,
+      "step": 4263
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5862549759040205,
+      "learning_rate": 8.545549487970577e-06,
+      "loss": 0.5268,
+      "step": 4264
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7393124458187841,
+      "learning_rate": 8.544822043165326e-06,
+      "loss": 0.5998,
+      "step": 4265
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.660501968995592,
+      "learning_rate": 8.544094447468328e-06,
+      "loss": 0.594,
+      "step": 4266
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.694251272036348,
+      "learning_rate": 8.543366700910555e-06,
+      "loss": 0.5976,
+      "step": 4267
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7277816593245259,
+      "learning_rate": 8.542638803522985e-06,
+      "loss": 0.5914,
+      "step": 4268
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7565444795806341,
+      "learning_rate": 8.5419107553366e-06,
+      "loss": 0.6361,
+      "step": 4269
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8124572372090746,
+      "learning_rate": 8.541182556382393e-06,
+      "loss": 0.6094,
+      "step": 4270
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5984985772708591,
+      "learning_rate": 8.540454206691362e-06,
+      "loss": 0.5356,
+      "step": 4271
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6298035543755568,
+      "learning_rate": 8.539725706294509e-06,
+      "loss": 0.5158,
+      "step": 4272
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8228448548212153,
+      "learning_rate": 8.538997055222842e-06,
+      "loss": 0.5704,
+      "step": 4273
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8257302554424572,
+      "learning_rate": 8.53826825350738e-06,
+      "loss": 0.6344,
+      "step": 4274
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7226723426272715,
+      "learning_rate": 8.537539301179145e-06,
+      "loss": 0.6658,
+      "step": 4275
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8743980343808205,
+      "learning_rate": 8.536810198269168e-06,
+      "loss": 0.5995,
+      "step": 4276
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7517533146836193,
+      "learning_rate": 8.536080944808483e-06,
+      "loss": 0.6374,
+      "step": 4277
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6061560152620385,
+      "learning_rate": 8.53535154082813e-06,
+      "loss": 0.5455,
+      "step": 4278
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6457503016339637,
+      "learning_rate": 8.53462198635916e-06,
+      "loss": 0.5165,
+      "step": 4279
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6406034037273896,
+      "learning_rate": 8.533892281432627e-06,
+      "loss": 0.5416,
+      "step": 4280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6012821386061982,
+      "learning_rate": 8.533162426079591e-06,
+      "loss": 0.5397,
+      "step": 4281
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6000210809009845,
+      "learning_rate": 8.532432420331122e-06,
+      "loss": 0.5231,
+      "step": 4282
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8292004198863734,
+      "learning_rate": 8.531702264218295e-06,
+      "loss": 0.6071,
+      "step": 4283
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8269181989750112,
+      "learning_rate": 8.530971957772186e-06,
+      "loss": 0.6241,
+      "step": 4284
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7960326046520969,
+      "learning_rate": 8.530241501023885e-06,
+      "loss": 0.673,
+      "step": 4285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6209979527108801,
+      "learning_rate": 8.529510894004484e-06,
+      "loss": 0.5193,
+      "step": 4286
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8199632399794136,
+      "learning_rate": 8.528780136745083e-06,
+      "loss": 0.5951,
+      "step": 4287
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7524330032267278,
+      "learning_rate": 8.528049229276787e-06,
+      "loss": 0.6162,
+      "step": 4288
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7915587553085761,
+      "learning_rate": 8.52731817163071e-06,
+      "loss": 0.6037,
+      "step": 4289
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8825922227382288,
+      "learning_rate": 8.52658696383797e-06,
+      "loss": 0.6703,
+      "step": 4290
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8712034370811429,
+      "learning_rate": 8.525855605929694e-06,
+      "loss": 0.5956,
+      "step": 4291
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7057093369300327,
+      "learning_rate": 8.52512409793701e-06,
+      "loss": 0.4895,
+      "step": 4292
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7541501445141401,
+      "learning_rate": 8.524392439891058e-06,
+      "loss": 0.6444,
+      "step": 4293
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7671029700881294,
+      "learning_rate": 8.523660631822983e-06,
+      "loss": 0.5962,
+      "step": 4294
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0784969699630877,
+      "learning_rate": 8.522928673763934e-06,
+      "loss": 0.6503,
+      "step": 4295
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5629563927965483,
+      "learning_rate": 8.52219656574507e-06,
+      "loss": 0.4987,
+      "step": 4296
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9040638419303705,
+      "learning_rate": 8.521464307797553e-06,
+      "loss": 0.6195,
+      "step": 4297
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5859049259778425,
+      "learning_rate": 8.520731899952553e-06,
+      "loss": 0.4963,
+      "step": 4298
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8656947498642173,
+      "learning_rate": 8.519999342241247e-06,
+      "loss": 0.5696,
+      "step": 4299
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7548629521197194,
+      "learning_rate": 8.519266634694817e-06,
+      "loss": 0.6493,
+      "step": 4300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5617877931928973,
+      "learning_rate": 8.518533777344453e-06,
+      "loss": 0.4903,
+      "step": 4301
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8194925530236241,
+      "learning_rate": 8.51780077022135e-06,
+      "loss": 0.6523,
+      "step": 4302
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7949077097631095,
+      "learning_rate": 8.517067613356708e-06,
+      "loss": 0.6174,
+      "step": 4303
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6836000824920909,
+      "learning_rate": 8.516334306781737e-06,
+      "loss": 0.4949,
+      "step": 4304
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6349075682784242,
+      "learning_rate": 8.515600850527653e-06,
+      "loss": 0.4851,
+      "step": 4305
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6277272316069593,
+      "learning_rate": 8.514867244625673e-06,
+      "loss": 0.5361,
+      "step": 4306
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7849166812396969,
+      "learning_rate": 8.514133489107028e-06,
+      "loss": 0.6509,
+      "step": 4307
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7560839783702589,
+      "learning_rate": 8.513399584002948e-06,
+      "loss": 0.6344,
+      "step": 4308
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5989992762340549,
+      "learning_rate": 8.512665529344675e-06,
+      "loss": 0.5297,
+      "step": 4309
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7960445312315848,
+      "learning_rate": 8.511931325163457e-06,
+      "loss": 0.6894,
+      "step": 4310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6863474292387314,
+      "learning_rate": 8.511196971490546e-06,
+      "loss": 0.5417,
+      "step": 4311
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7034927102973783,
+      "learning_rate": 8.510462468357197e-06,
+      "loss": 0.5478,
+      "step": 4312
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7326193985658955,
+      "learning_rate": 8.50972781579468e-06,
+      "loss": 0.643,
+      "step": 4313
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.630221023403941,
+      "learning_rate": 8.508993013834266e-06,
+      "loss": 0.5863,
+      "step": 4314
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.854629603371839,
+      "learning_rate": 8.508258062507231e-06,
+      "loss": 0.7465,
+      "step": 4315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.744227787448643,
+      "learning_rate": 8.507522961844861e-06,
+      "loss": 0.6361,
+      "step": 4316
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8040693037841271,
+      "learning_rate": 8.50678771187845e-06,
+      "loss": 0.6694,
+      "step": 4317
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8317219420549856,
+      "learning_rate": 8.506052312639288e-06,
+      "loss": 0.6361,
+      "step": 4318
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6828791795082905,
+      "learning_rate": 8.505316764158685e-06,
+      "loss": 0.5857,
+      "step": 4319
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6833364312519926,
+      "learning_rate": 8.504581066467949e-06,
+      "loss": 0.6378,
+      "step": 4320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.604475259089991,
+      "learning_rate": 8.503845219598395e-06,
+      "loss": 0.5891,
+      "step": 4321
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8001010618198438,
+      "learning_rate": 8.503109223581348e-06,
+      "loss": 0.6729,
+      "step": 4322
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7029522662113565,
+      "learning_rate": 8.502373078448136e-06,
+      "loss": 0.5837,
+      "step": 4323
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6558529097763018,
+      "learning_rate": 8.501636784230091e-06,
+      "loss": 0.6184,
+      "step": 4324
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7121771072838233,
+      "learning_rate": 8.50090034095856e-06,
+      "loss": 0.5648,
+      "step": 4325
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7918856170257168,
+      "learning_rate": 8.50016374866489e-06,
+      "loss": 0.6181,
+      "step": 4326
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.744418829081397,
+      "learning_rate": 8.499427007380433e-06,
+      "loss": 0.6777,
+      "step": 4327
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8242404096462485,
+      "learning_rate": 8.49869011713655e-06,
+      "loss": 0.5594,
+      "step": 4328
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8580536800368124,
+      "learning_rate": 8.49795307796461e-06,
+      "loss": 0.6931,
+      "step": 4329
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.757476645322607,
+      "learning_rate": 8.497215889895984e-06,
+      "loss": 0.6423,
+      "step": 4330
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5462465331305486,
+      "learning_rate": 8.496478552962053e-06,
+      "loss": 0.557,
+      "step": 4331
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7154005967583038,
+      "learning_rate": 8.495741067194203e-06,
+      "loss": 0.6269,
+      "step": 4332
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7064131698022073,
+      "learning_rate": 8.495003432623828e-06,
+      "loss": 0.6145,
+      "step": 4333
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.578074893201352,
+      "learning_rate": 8.494265649282325e-06,
+      "loss": 0.5458,
+      "step": 4334
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5167112198028618,
+      "learning_rate": 8.4935277172011e-06,
+      "loss": 0.5034,
+      "step": 4335
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7197536131515607,
+      "learning_rate": 8.492789636411562e-06,
+      "loss": 0.6008,
+      "step": 4336
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8414455711739971,
+      "learning_rate": 8.492051406945132e-06,
+      "loss": 0.6627,
+      "step": 4337
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8546785748275249,
+      "learning_rate": 8.491313028833232e-06,
+      "loss": 0.692,
+      "step": 4338
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5943690480985239,
+      "learning_rate": 8.490574502107294e-06,
+      "loss": 0.5173,
+      "step": 4339
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7680057826779969,
+      "learning_rate": 8.489835826798752e-06,
+      "loss": 0.6163,
+      "step": 4340
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5888222986238567,
+      "learning_rate": 8.489097002939052e-06,
+      "loss": 0.504,
+      "step": 4341
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6894175108984276,
+      "learning_rate": 8.488358030559643e-06,
+      "loss": 0.6367,
+      "step": 4342
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.740876167443951,
+      "learning_rate": 8.487618909691977e-06,
+      "loss": 0.6399,
+      "step": 4343
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7171090480709094,
+      "learning_rate": 8.48687964036752e-06,
+      "loss": 0.5975,
+      "step": 4344
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7724993531563258,
+      "learning_rate": 8.48614022261774e-06,
+      "loss": 0.6287,
+      "step": 4345
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7967773280677185,
+      "learning_rate": 8.48540065647411e-06,
+      "loss": 0.5835,
+      "step": 4346
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6120062632400328,
+      "learning_rate": 8.484660941968113e-06,
+      "loss": 0.5514,
+      "step": 4347
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6610184084121873,
+      "learning_rate": 8.483921079131233e-06,
+      "loss": 0.5538,
+      "step": 4348
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6147261331619809,
+      "learning_rate": 8.48318106799497e-06,
+      "loss": 0.4788,
+      "step": 4349
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7151182122048538,
+      "learning_rate": 8.482440908590816e-06,
+      "loss": 0.571,
+      "step": 4350
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7994778922305898,
+      "learning_rate": 8.48170060095028e-06,
+      "loss": 0.5999,
+      "step": 4351
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9544782281327197,
+      "learning_rate": 8.480960145104877e-06,
+      "loss": 0.6318,
+      "step": 4352
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7274607856506495,
+      "learning_rate": 8.480219541086125e-06,
+      "loss": 0.619,
+      "step": 4353
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8166137352707963,
+      "learning_rate": 8.479478788925547e-06,
+      "loss": 0.6274,
+      "step": 4354
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7180740681173754,
+      "learning_rate": 8.478737888654676e-06,
+      "loss": 0.5898,
+      "step": 4355
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5609736243588401,
+      "learning_rate": 8.477996840305049e-06,
+      "loss": 0.4804,
+      "step": 4356
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5991913989693161,
+      "learning_rate": 8.47725564390821e-06,
+      "loss": 0.5141,
+      "step": 4357
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.894912278276884,
+      "learning_rate": 8.476514299495713e-06,
+      "loss": 0.6199,
+      "step": 4358
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.647204864992118,
+      "learning_rate": 8.47577280709911e-06,
+      "loss": 0.5633,
+      "step": 4359
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7095036023821971,
+      "learning_rate": 8.475031166749965e-06,
+      "loss": 0.5271,
+      "step": 4360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6218753907478503,
+      "learning_rate": 8.47428937847985e-06,
+      "loss": 0.5431,
+      "step": 4361
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6700308475198772,
+      "learning_rate": 8.473547442320335e-06,
+      "loss": 0.5286,
+      "step": 4362
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7376301008276802,
+      "learning_rate": 8.472805358303007e-06,
+      "loss": 0.6417,
+      "step": 4363
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7932546849779789,
+      "learning_rate": 8.472063126459453e-06,
+      "loss": 0.5774,
+      "step": 4364
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7437735077644912,
+      "learning_rate": 8.471320746821265e-06,
+      "loss": 0.5766,
+      "step": 4365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8088037529803853,
+      "learning_rate": 8.470578219420048e-06,
+      "loss": 0.6105,
+      "step": 4366
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0850284063236828,
+      "learning_rate": 8.469835544287404e-06,
+      "loss": 0.6306,
+      "step": 4367
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6899848915368455,
+      "learning_rate": 8.46909272145495e-06,
+      "loss": 0.6038,
+      "step": 4368
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6172932585039929,
+      "learning_rate": 8.468349750954306e-06,
+      "loss": 0.5382,
+      "step": 4369
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8093365987408733,
+      "learning_rate": 8.467606632817093e-06,
+      "loss": 0.661,
+      "step": 4370
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8171873486748183,
+      "learning_rate": 8.466863367074949e-06,
+      "loss": 0.6363,
+      "step": 4371
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7271344873956909,
+      "learning_rate": 8.46611995375951e-06,
+      "loss": 0.4952,
+      "step": 4372
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.615570178408858,
+      "learning_rate": 8.46537639290242e-06,
+      "loss": 0.5651,
+      "step": 4373
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6726183851577787,
+      "learning_rate": 8.46463268453533e-06,
+      "loss": 0.5349,
+      "step": 4374
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7369120502687199,
+      "learning_rate": 8.4638888286899e-06,
+      "loss": 0.6024,
+      "step": 4375
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0593756232299247,
+      "learning_rate": 8.463144825397792e-06,
+      "loss": 0.611,
+      "step": 4376
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6251734736496313,
+      "learning_rate": 8.462400674690673e-06,
+      "loss": 0.5461,
+      "step": 4377
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7709423700590702,
+      "learning_rate": 8.461656376600222e-06,
+      "loss": 0.665,
+      "step": 4378
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7453921369022694,
+      "learning_rate": 8.460911931158122e-06,
+      "loss": 0.5916,
+      "step": 4379
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.785242195667744,
+      "learning_rate": 8.46016733839606e-06,
+      "loss": 0.5633,
+      "step": 4380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8122768105789598,
+      "learning_rate": 8.45942259834573e-06,
+      "loss": 0.6906,
+      "step": 4381
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9553098162617438,
+      "learning_rate": 8.458677711038836e-06,
+      "loss": 0.6909,
+      "step": 4382
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7384939551402149,
+      "learning_rate": 8.457932676507085e-06,
+      "loss": 0.5796,
+      "step": 4383
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8338925407618774,
+      "learning_rate": 8.457187494782189e-06,
+      "loss": 0.6332,
+      "step": 4384
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6910187485770999,
+      "learning_rate": 8.456442165895869e-06,
+      "loss": 0.6281,
+      "step": 4385
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5985086222563284,
+      "learning_rate": 8.455696689879852e-06,
+      "loss": 0.5164,
+      "step": 4386
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7031678249461256,
+      "learning_rate": 8.454951066765868e-06,
+      "loss": 0.5555,
+      "step": 4387
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8621667979416765,
+      "learning_rate": 8.45420529658566e-06,
+      "loss": 0.6076,
+      "step": 4388
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9141387606178659,
+      "learning_rate": 8.453459379370969e-06,
+      "loss": 0.6112,
+      "step": 4389
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6266554619922422,
+      "learning_rate": 8.45271331515355e-06,
+      "loss": 0.527,
+      "step": 4390
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8093966545388899,
+      "learning_rate": 8.451967103965155e-06,
+      "loss": 0.65,
+      "step": 4391
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9465313148657902,
+      "learning_rate": 8.451220745837553e-06,
+      "loss": 0.6203,
+      "step": 4392
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8207052142712354,
+      "learning_rate": 8.45047424080251e-06,
+      "loss": 0.6107,
+      "step": 4393
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6106219759665608,
+      "learning_rate": 8.44972758889181e-06,
+      "loss": 0.5803,
+      "step": 4394
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7100279341225781,
+      "learning_rate": 8.448980790137226e-06,
+      "loss": 0.5977,
+      "step": 4395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.625937153542458,
+      "learning_rate": 8.448233844570553e-06,
+      "loss": 0.5184,
+      "step": 4396
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7864800476436626,
+      "learning_rate": 8.447486752223584e-06,
+      "loss": 0.5955,
+      "step": 4397
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7887189384852409,
+      "learning_rate": 8.44673951312812e-06,
+      "loss": 0.6166,
+      "step": 4398
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7566261208448904,
+      "learning_rate": 8.445992127315969e-06,
+      "loss": 0.5728,
+      "step": 4399
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5423403336054106,
+      "learning_rate": 8.445244594818945e-06,
+      "loss": 0.4905,
+      "step": 4400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.621849865909122,
+      "learning_rate": 8.44449691566887e-06,
+      "loss": 0.5985,
+      "step": 4401
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7771931513536706,
+      "learning_rate": 8.443749089897564e-06,
+      "loss": 0.598,
+      "step": 4402
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8557942555061883,
+      "learning_rate": 8.443001117536868e-06,
+      "loss": 0.6443,
+      "step": 4403
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6672955356737872,
+      "learning_rate": 8.442252998618615e-06,
+      "loss": 0.5755,
+      "step": 4404
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.580889998246896,
+      "learning_rate": 8.441504733174653e-06,
+      "loss": 0.5515,
+      "step": 4405
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7432901191602979,
+      "learning_rate": 8.440756321236829e-06,
+      "loss": 0.591,
+      "step": 4406
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9234935615780759,
+      "learning_rate": 8.440007762837006e-06,
+      "loss": 0.6407,
+      "step": 4407
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.649441462722885,
+      "learning_rate": 8.439259058007046e-06,
+      "loss": 0.5522,
+      "step": 4408
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5782242670876485,
+      "learning_rate": 8.438510206778816e-06,
+      "loss": 0.5429,
+      "step": 4409
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.623231150960479,
+      "learning_rate": 8.437761209184196e-06,
+      "loss": 0.5516,
+      "step": 4410
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8951910480939944,
+      "learning_rate": 8.437012065255066e-06,
+      "loss": 0.6411,
+      "step": 4411
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5606392635611586,
+      "learning_rate": 8.436262775023314e-06,
+      "loss": 0.513,
+      "step": 4412
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.533246558256732,
+      "learning_rate": 8.435513338520839e-06,
+      "loss": 0.54,
+      "step": 4413
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6118557654265043,
+      "learning_rate": 8.434763755779538e-06,
+      "loss": 0.5256,
+      "step": 4414
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0038463817566687,
+      "learning_rate": 8.434014026831321e-06,
+      "loss": 0.5858,
+      "step": 4415
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5914753578370091,
+      "learning_rate": 8.4332641517081e-06,
+      "loss": 0.4763,
+      "step": 4416
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7647039799522053,
+      "learning_rate": 8.432514130441794e-06,
+      "loss": 0.6039,
+      "step": 4417
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7876606790890524,
+      "learning_rate": 8.431763963064331e-06,
+      "loss": 0.667,
+      "step": 4418
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6997141913370188,
+      "learning_rate": 8.431013649607643e-06,
+      "loss": 0.6125,
+      "step": 4419
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7125075326371043,
+      "learning_rate": 8.430263190103667e-06,
+      "loss": 0.5906,
+      "step": 4420
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9033565393958323,
+      "learning_rate": 8.42951258458435e-06,
+      "loss": 0.5742,
+      "step": 4421
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.618547274382211,
+      "learning_rate": 8.42876183308164e-06,
+      "loss": 0.5284,
+      "step": 4422
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.641059564173985,
+      "learning_rate": 8.428010935627496e-06,
+      "loss": 0.551,
+      "step": 4423
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6386043167072742,
+      "learning_rate": 8.42725989225388e-06,
+      "loss": 0.5118,
+      "step": 4424
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6912236302258694,
+      "learning_rate": 8.426508702992763e-06,
+      "loss": 0.5888,
+      "step": 4425
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8647033230811324,
+      "learning_rate": 8.425757367876122e-06,
+      "loss": 0.6452,
+      "step": 4426
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.608122830630269,
+      "learning_rate": 8.425005886935936e-06,
+      "loss": 0.5214,
+      "step": 4427
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5668157024818089,
+      "learning_rate": 8.424254260204192e-06,
+      "loss": 0.5299,
+      "step": 4428
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7158474583360988,
+      "learning_rate": 8.423502487712888e-06,
+      "loss": 0.5725,
+      "step": 4429
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7657430339608319,
+      "learning_rate": 8.422750569494023e-06,
+      "loss": 0.5384,
+      "step": 4430
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7803553538455631,
+      "learning_rate": 8.421998505579603e-06,
+      "loss": 0.6445,
+      "step": 4431
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6182084913311109,
+      "learning_rate": 8.421246296001644e-06,
+      "loss": 0.5148,
+      "step": 4432
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8661938537133296,
+      "learning_rate": 8.42049394079216e-06,
+      "loss": 0.5838,
+      "step": 4433
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5759663788211165,
+      "learning_rate": 8.419741439983181e-06,
+      "loss": 0.47,
+      "step": 4434
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7058287456561901,
+      "learning_rate": 8.418988793606738e-06,
+      "loss": 0.6008,
+      "step": 4435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7666808515140944,
+      "learning_rate": 8.418236001694864e-06,
+      "loss": 0.61,
+      "step": 4436
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5853176142649072,
+      "learning_rate": 8.41748306427961e-06,
+      "loss": 0.5359,
+      "step": 4437
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5862092657965132,
+      "learning_rate": 8.416729981393021e-06,
+      "loss": 0.456,
+      "step": 4438
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5724659081903708,
+      "learning_rate": 8.415976753067154e-06,
+      "loss": 0.5349,
+      "step": 4439
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6833715600826622,
+      "learning_rate": 8.415223379334075e-06,
+      "loss": 0.5165,
+      "step": 4440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7931207098380295,
+      "learning_rate": 8.414469860225849e-06,
+      "loss": 0.6231,
+      "step": 4441
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6813979269344471,
+      "learning_rate": 8.413716195774553e-06,
+      "loss": 0.533,
+      "step": 4442
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8431388638660972,
+      "learning_rate": 8.412962386012264e-06,
+      "loss": 0.652,
+      "step": 4443
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7565321624882378,
+      "learning_rate": 8.412208430971076e-06,
+      "loss": 0.5863,
+      "step": 4444
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7096067574235724,
+      "learning_rate": 8.411454330683077e-06,
+      "loss": 0.5656,
+      "step": 4445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6383764971101797,
+      "learning_rate": 8.41070008518037e-06,
+      "loss": 0.5137,
+      "step": 4446
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7497706062021322,
+      "learning_rate": 8.40994569449506e-06,
+      "loss": 0.6202,
+      "step": 4447
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6312542897477779,
+      "learning_rate": 8.409191158659259e-06,
+      "loss": 0.5386,
+      "step": 4448
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.653863516156257,
+      "learning_rate": 8.40843647770508e-06,
+      "loss": 0.5385,
+      "step": 4449
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6156252252004295,
+      "learning_rate": 8.407681651664654e-06,
+      "loss": 0.5179,
+      "step": 4450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7142752209597169,
+      "learning_rate": 8.406926680570111e-06,
+      "loss": 0.5873,
+      "step": 4451
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0733664514333419,
+      "learning_rate": 8.406171564453588e-06,
+      "loss": 0.6702,
+      "step": 4452
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5710862247044152,
+      "learning_rate": 8.405416303347223e-06,
+      "loss": 0.5509,
+      "step": 4453
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7607990110698681,
+      "learning_rate": 8.404660897283168e-06,
+      "loss": 0.5989,
+      "step": 4454
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8445745815776041,
+      "learning_rate": 8.403905346293579e-06,
+      "loss": 0.6713,
+      "step": 4455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7937089993604419,
+      "learning_rate": 8.403149650410617e-06,
+      "loss": 0.6041,
+      "step": 4456
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7770958593130254,
+      "learning_rate": 8.402393809666448e-06,
+      "loss": 0.6032,
+      "step": 4457
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7400065671466064,
+      "learning_rate": 8.40163782409325e-06,
+      "loss": 0.6354,
+      "step": 4458
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7432366006126222,
+      "learning_rate": 8.400881693723197e-06,
+      "loss": 0.6646,
+      "step": 4459
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7735783971083421,
+      "learning_rate": 8.400125418588478e-06,
+      "loss": 0.5998,
+      "step": 4460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7964233894410391,
+      "learning_rate": 8.399368998721287e-06,
+      "loss": 0.6373,
+      "step": 4461
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6999861157095514,
+      "learning_rate": 8.398612434153819e-06,
+      "loss": 0.6027,
+      "step": 4462
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6412778795231933,
+      "learning_rate": 8.39785572491828e-06,
+      "loss": 0.5091,
+      "step": 4463
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7456919161309303,
+      "learning_rate": 8.397098871046883e-06,
+      "loss": 0.6147,
+      "step": 4464
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5393626049617862,
+      "learning_rate": 8.396341872571841e-06,
+      "loss": 0.5005,
+      "step": 4465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8533845519661352,
+      "learning_rate": 8.395584729525379e-06,
+      "loss": 0.5932,
+      "step": 4466
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6321572046611634,
+      "learning_rate": 8.394827441939725e-06,
+      "loss": 0.5685,
+      "step": 4467
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6953148304892535,
+      "learning_rate": 8.394070009847115e-06,
+      "loss": 0.6013,
+      "step": 4468
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9975105974700247,
+      "learning_rate": 8.393312433279791e-06,
+      "loss": 0.6189,
+      "step": 4469
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6276522101524512,
+      "learning_rate": 8.392554712270001e-06,
+      "loss": 0.5636,
+      "step": 4470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7611873920784217,
+      "learning_rate": 8.391796846849997e-06,
+      "loss": 0.6058,
+      "step": 4471
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6066135351030368,
+      "learning_rate": 8.391038837052042e-06,
+      "loss": 0.5753,
+      "step": 4472
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5548566589434175,
+      "learning_rate": 8.390280682908398e-06,
+      "loss": 0.5037,
+      "step": 4473
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6370250625053674,
+      "learning_rate": 8.389522384451338e-06,
+      "loss": 0.4581,
+      "step": 4474
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0508030236374957,
+      "learning_rate": 8.388763941713142e-06,
+      "loss": 0.6039,
+      "step": 4475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5595806131880281,
+      "learning_rate": 8.388005354726096e-06,
+      "loss": 0.4803,
+      "step": 4476
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.833762981885313,
+      "learning_rate": 8.387246623522487e-06,
+      "loss": 0.6109,
+      "step": 4477
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7650249292107758,
+      "learning_rate": 8.386487748134615e-06,
+      "loss": 0.5198,
+      "step": 4478
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2155262611753492,
+      "learning_rate": 8.38572872859478e-06,
+      "loss": 0.6279,
+      "step": 4479
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6689505057411278,
+      "learning_rate": 8.38496956493529e-06,
+      "loss": 0.5565,
+      "step": 4480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8908804902154505,
+      "learning_rate": 8.384210257188465e-06,
+      "loss": 0.6179,
+      "step": 4481
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8230118279260698,
+      "learning_rate": 8.383450805386624e-06,
+      "loss": 0.6429,
+      "step": 4482
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.576327363748139,
+      "learning_rate": 8.382691209562093e-06,
+      "loss": 0.4824,
+      "step": 4483
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.608708951314828,
+      "learning_rate": 8.381931469747208e-06,
+      "loss": 0.539,
+      "step": 4484
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8149030142574907,
+      "learning_rate": 8.381171585974306e-06,
+      "loss": 0.569,
+      "step": 4485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8041259526101094,
+      "learning_rate": 8.380411558275735e-06,
+      "loss": 0.6296,
+      "step": 4486
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7740474608758181,
+      "learning_rate": 8.379651386683845e-06,
+      "loss": 0.5857,
+      "step": 4487
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6021501709260586,
+      "learning_rate": 8.378891071230998e-06,
+      "loss": 0.4878,
+      "step": 4488
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5518869268865115,
+      "learning_rate": 8.378130611949553e-06,
+      "loss": 0.5544,
+      "step": 4489
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8623381337388066,
+      "learning_rate": 8.377370008871884e-06,
+      "loss": 0.6404,
+      "step": 4490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6187203560978535,
+      "learning_rate": 8.376609262030366e-06,
+      "loss": 0.5227,
+      "step": 4491
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6501240335984548,
+      "learning_rate": 8.375848371457383e-06,
+      "loss": 0.564,
+      "step": 4492
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8229738642659958,
+      "learning_rate": 8.37508733718532e-06,
+      "loss": 0.6042,
+      "step": 4493
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7727942980664513,
+      "learning_rate": 8.374326159246575e-06,
+      "loss": 0.6231,
+      "step": 4494
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7508246557879681,
+      "learning_rate": 8.373564837673549e-06,
+      "loss": 0.5802,
+      "step": 4495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5813973478855676,
+      "learning_rate": 8.372803372498648e-06,
+      "loss": 0.4964,
+      "step": 4496
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7093755340221902,
+      "learning_rate": 8.372041763754287e-06,
+      "loss": 0.5232,
+      "step": 4497
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7251925775924358,
+      "learning_rate": 8.371280011472881e-06,
+      "loss": 0.5971,
+      "step": 4498
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5501750394360069,
+      "learning_rate": 8.370518115686858e-06,
+      "loss": 0.5746,
+      "step": 4499
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6571418849244468,
+      "learning_rate": 8.369756076428652e-06,
+      "loss": 0.5331,
+      "step": 4500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5686787720364862,
+      "learning_rate": 8.368993893730697e-06,
+      "loss": 0.5394,
+      "step": 4501
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5638748476950008,
+      "learning_rate": 8.368231567625437e-06,
+      "loss": 0.5315,
+      "step": 4502
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7717308841066829,
+      "learning_rate": 8.367469098145322e-06,
+      "loss": 0.6847,
+      "step": 4503
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8644609082076836,
+      "learning_rate": 8.366706485322808e-06,
+      "loss": 0.5843,
+      "step": 4504
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6412821334544353,
+      "learning_rate": 8.36594372919036e-06,
+      "loss": 0.5142,
+      "step": 4505
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5799824683265785,
+      "learning_rate": 8.365180829780442e-06,
+      "loss": 0.5174,
+      "step": 4506
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.568866374908323,
+      "learning_rate": 8.36441778712553e-06,
+      "loss": 0.5253,
+      "step": 4507
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7030245093748975,
+      "learning_rate": 8.363654601258103e-06,
+      "loss": 0.5826,
+      "step": 4508
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.854340025941681,
+      "learning_rate": 8.362891272210647e-06,
+      "loss": 0.6624,
+      "step": 4509
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7255927073232108,
+      "learning_rate": 8.362127800015659e-06,
+      "loss": 0.5874,
+      "step": 4510
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5250171534714101,
+      "learning_rate": 8.361364184705633e-06,
+      "loss": 0.483,
+      "step": 4511
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7513306481367137,
+      "learning_rate": 8.360600426313074e-06,
+      "loss": 0.5445,
+      "step": 4512
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6615244418505919,
+      "learning_rate": 8.359836524870496e-06,
+      "loss": 0.5322,
+      "step": 4513
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7702991030071065,
+      "learning_rate": 8.359072480410412e-06,
+      "loss": 0.6721,
+      "step": 4514
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5494868261672948,
+      "learning_rate": 8.358308292965347e-06,
+      "loss": 0.5264,
+      "step": 4515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6066363350537363,
+      "learning_rate": 8.35754396256783e-06,
+      "loss": 0.5767,
+      "step": 4516
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7460154252299692,
+      "learning_rate": 8.356779489250395e-06,
+      "loss": 0.566,
+      "step": 4517
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.647224113721489,
+      "learning_rate": 8.356014873045584e-06,
+      "loss": 0.5135,
+      "step": 4518
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7487485443601356,
+      "learning_rate": 8.355250113985945e-06,
+      "loss": 0.6398,
+      "step": 4519
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7762880385238963,
+      "learning_rate": 8.35448521210403e-06,
+      "loss": 0.6106,
+      "step": 4520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7100629932358956,
+      "learning_rate": 8.353720167432397e-06,
+      "loss": 0.6517,
+      "step": 4521
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6947478186112446,
+      "learning_rate": 8.352954980003615e-06,
+      "loss": 0.528,
+      "step": 4522
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7519450452872166,
+      "learning_rate": 8.352189649850255e-06,
+      "loss": 0.5414,
+      "step": 4523
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7199672055423483,
+      "learning_rate": 8.351424177004893e-06,
+      "loss": 0.5983,
+      "step": 4524
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6369900256999003,
+      "learning_rate": 8.350658561500114e-06,
+      "loss": 0.4919,
+      "step": 4525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6968041918701267,
+      "learning_rate": 8.349892803368506e-06,
+      "loss": 0.6362,
+      "step": 4526
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6900397909829166,
+      "learning_rate": 8.349126902642667e-06,
+      "loss": 0.5738,
+      "step": 4527
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7267647190814117,
+      "learning_rate": 8.348360859355198e-06,
+      "loss": 0.6043,
+      "step": 4528
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.779604519713317,
+      "learning_rate": 8.347594673538705e-06,
+      "loss": 0.6071,
+      "step": 4529
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.743336975063652,
+      "learning_rate": 8.346828345225806e-06,
+      "loss": 0.6383,
+      "step": 4530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.544228793325572,
+      "learning_rate": 8.34606187444912e-06,
+      "loss": 0.4821,
+      "step": 4531
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7831383892292381,
+      "learning_rate": 8.34529526124127e-06,
+      "loss": 0.6049,
+      "step": 4532
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7421778872396227,
+      "learning_rate": 8.344528505634894e-06,
+      "loss": 0.6024,
+      "step": 4533
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8358407921835102,
+      "learning_rate": 8.343761607662625e-06,
+      "loss": 0.6336,
+      "step": 4534
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7708064486421488,
+      "learning_rate": 8.34299456735711e-06,
+      "loss": 0.622,
+      "step": 4535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6063444648625697,
+      "learning_rate": 8.342227384750998e-06,
+      "loss": 0.5267,
+      "step": 4536
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5994602788328899,
+      "learning_rate": 8.341460059876947e-06,
+      "loss": 0.5056,
+      "step": 4537
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.563371386041275,
+      "learning_rate": 8.34069259276762e-06,
+      "loss": 0.4971,
+      "step": 4538
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6961521146768033,
+      "learning_rate": 8.339924983455684e-06,
+      "loss": 0.5955,
+      "step": 4539
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7742789902056458,
+      "learning_rate": 8.339157231973815e-06,
+      "loss": 0.5895,
+      "step": 4540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6445128213410857,
+      "learning_rate": 8.338389338354693e-06,
+      "loss": 0.5341,
+      "step": 4541
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6121826947430126,
+      "learning_rate": 8.337621302631003e-06,
+      "loss": 0.5428,
+      "step": 4542
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7345697368035174,
+      "learning_rate": 8.33685312483544e-06,
+      "loss": 0.5354,
+      "step": 4543
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8811001564261831,
+      "learning_rate": 8.336084805000703e-06,
+      "loss": 0.6578,
+      "step": 4544
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7278108537264646,
+      "learning_rate": 8.335316343159498e-06,
+      "loss": 0.57,
+      "step": 4545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6591807933465704,
+      "learning_rate": 8.334547739344534e-06,
+      "loss": 0.5237,
+      "step": 4546
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6241909819167553,
+      "learning_rate": 8.333778993588529e-06,
+      "loss": 0.5383,
+      "step": 4547
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5406844452349486,
+      "learning_rate": 8.333010105924201e-06,
+      "loss": 0.4765,
+      "step": 4548
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5974736064406457,
+      "learning_rate": 8.33224107638429e-06,
+      "loss": 0.528,
+      "step": 4549
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6932640265580899,
+      "learning_rate": 8.331471905001521e-06,
+      "loss": 0.5598,
+      "step": 4550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6459708961663617,
+      "learning_rate": 8.330702591808639e-06,
+      "loss": 0.4623,
+      "step": 4551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6357239836619667,
+      "learning_rate": 8.329933136838391e-06,
+      "loss": 0.5339,
+      "step": 4552
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8551503830518864,
+      "learning_rate": 8.32916354012353e-06,
+      "loss": 0.6803,
+      "step": 4553
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5869725208040968,
+      "learning_rate": 8.328393801696817e-06,
+      "loss": 0.4797,
+      "step": 4554
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6419183139505903,
+      "learning_rate": 8.327623921591014e-06,
+      "loss": 0.5271,
+      "step": 4555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8884931704165502,
+      "learning_rate": 8.326853899838895e-06,
+      "loss": 0.6423,
+      "step": 4556
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7618765452436912,
+      "learning_rate": 8.326083736473238e-06,
+      "loss": 0.6033,
+      "step": 4557
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.821669465018436,
+      "learning_rate": 8.325313431526824e-06,
+      "loss": 0.6335,
+      "step": 4558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5591507822069133,
+      "learning_rate": 8.324542985032444e-06,
+      "loss": 0.5383,
+      "step": 4559
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5881586167867418,
+      "learning_rate": 8.32377239702289e-06,
+      "loss": 0.544,
+      "step": 4560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9188463807647256,
+      "learning_rate": 8.32300166753097e-06,
+      "loss": 0.5166,
+      "step": 4561
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8610061131662988,
+      "learning_rate": 8.322230796589486e-06,
+      "loss": 0.595,
+      "step": 4562
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9617603272577288,
+      "learning_rate": 8.321459784231254e-06,
+      "loss": 0.6348,
+      "step": 4563
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7006695571472142,
+      "learning_rate": 8.320688630489093e-06,
+      "loss": 0.5687,
+      "step": 4564
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7845886099944425,
+      "learning_rate": 8.319917335395827e-06,
+      "loss": 0.5882,
+      "step": 4565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7221957013093857,
+      "learning_rate": 8.319145898984291e-06,
+      "loss": 0.5141,
+      "step": 4566
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.001814019951614,
+      "learning_rate": 8.318374321287319e-06,
+      "loss": 0.6619,
+      "step": 4567
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5920898880186863,
+      "learning_rate": 8.317602602337755e-06,
+      "loss": 0.5261,
+      "step": 4568
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8901935217083875,
+      "learning_rate": 8.316830742168452e-06,
+      "loss": 0.69,
+      "step": 4569
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.737862041467974,
+      "learning_rate": 8.316058740812263e-06,
+      "loss": 0.5711,
+      "step": 4570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6741381417129442,
+      "learning_rate": 8.31528659830205e-06,
+      "loss": 0.5838,
+      "step": 4571
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6252054561230648,
+      "learning_rate": 8.314514314670681e-06,
+      "loss": 0.5187,
+      "step": 4572
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6362798972986528,
+      "learning_rate": 8.313741889951028e-06,
+      "loss": 0.5418,
+      "step": 4573
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7232528490922306,
+      "learning_rate": 8.312969324175972e-06,
+      "loss": 0.5919,
+      "step": 4574
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8233850690240231,
+      "learning_rate": 8.312196617378399e-06,
+      "loss": 0.6409,
+      "step": 4575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6412296015838452,
+      "learning_rate": 8.311423769591201e-06,
+      "loss": 0.5617,
+      "step": 4576
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8807926764936764,
+      "learning_rate": 8.310650780847275e-06,
+      "loss": 0.6779,
+      "step": 4577
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6088078919307018,
+      "learning_rate": 8.309877651179523e-06,
+      "loss": 0.5354,
+      "step": 4578
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7457311656456699,
+      "learning_rate": 8.309104380620857e-06,
+      "loss": 0.5341,
+      "step": 4579
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5331643572616342,
+      "learning_rate": 8.308330969204192e-06,
+      "loss": 0.6517,
+      "step": 4580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6157849670100451,
+      "learning_rate": 8.30755741696245e-06,
+      "loss": 0.5307,
+      "step": 4581
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8538888377847706,
+      "learning_rate": 8.306783723928558e-06,
+      "loss": 0.6179,
+      "step": 4582
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5696347142685619,
+      "learning_rate": 8.30600989013545e-06,
+      "loss": 0.5387,
+      "step": 4583
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7679757721676566,
+      "learning_rate": 8.305235915616065e-06,
+      "loss": 0.6794,
+      "step": 4584
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6822017456418559,
+      "learning_rate": 8.30446180040335e-06,
+      "loss": 0.5187,
+      "step": 4585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6171342767274044,
+      "learning_rate": 8.303687544530254e-06,
+      "loss": 0.5268,
+      "step": 4586
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5780266415667538,
+      "learning_rate": 8.302913148029738e-06,
+      "loss": 0.528,
+      "step": 4587
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6674256120796203,
+      "learning_rate": 8.302138610934762e-06,
+      "loss": 0.5698,
+      "step": 4588
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8206296189598848,
+      "learning_rate": 8.3013639332783e-06,
+      "loss": 0.6536,
+      "step": 4589
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7877189165111409,
+      "learning_rate": 8.300589115093324e-06,
+      "loss": 0.635,
+      "step": 4590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5616090193575192,
+      "learning_rate": 8.299814156412816e-06,
+      "loss": 0.5161,
+      "step": 4591
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5988902365227176,
+      "learning_rate": 8.299039057269764e-06,
+      "loss": 0.5687,
+      "step": 4592
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6960419334930371,
+      "learning_rate": 8.298263817697162e-06,
+      "loss": 0.5795,
+      "step": 4593
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5707883144842965,
+      "learning_rate": 8.297488437728008e-06,
+      "loss": 0.5011,
+      "step": 4594
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6575111612549288,
+      "learning_rate": 8.296712917395311e-06,
+      "loss": 0.5649,
+      "step": 4595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.745300592831123,
+      "learning_rate": 8.295937256732077e-06,
+      "loss": 0.6651,
+      "step": 4596
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6442942869675553,
+      "learning_rate": 8.295161455771327e-06,
+      "loss": 0.5579,
+      "step": 4597
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6301425841386397,
+      "learning_rate": 8.294385514546083e-06,
+      "loss": 0.5346,
+      "step": 4598
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5985101956372935,
+      "learning_rate": 8.293609433089378e-06,
+      "loss": 0.5695,
+      "step": 4599
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5838525089912457,
+      "learning_rate": 8.292833211434243e-06,
+      "loss": 0.5491,
+      "step": 4600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.608297559285062,
+      "learning_rate": 8.29205684961372e-06,
+      "loss": 0.5415,
+      "step": 4601
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5850325101804145,
+      "learning_rate": 8.291280347660856e-06,
+      "loss": 0.53,
+      "step": 4602
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6541675043155815,
+      "learning_rate": 8.290503705608707e-06,
+      "loss": 0.5075,
+      "step": 4603
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.732558482290716,
+      "learning_rate": 8.28972692349033e-06,
+      "loss": 0.6011,
+      "step": 4604
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6349321427691408,
+      "learning_rate": 8.288950001338788e-06,
+      "loss": 0.5052,
+      "step": 4605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7066040299986205,
+      "learning_rate": 8.288172939187155e-06,
+      "loss": 0.5979,
+      "step": 4606
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6263607674648934,
+      "learning_rate": 8.287395737068509e-06,
+      "loss": 0.5699,
+      "step": 4607
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7576194191057484,
+      "learning_rate": 8.286618395015931e-06,
+      "loss": 0.5935,
+      "step": 4608
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6930537273041669,
+      "learning_rate": 8.28584091306251e-06,
+      "loss": 0.5415,
+      "step": 4609
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8121738648037318,
+      "learning_rate": 8.285063291241342e-06,
+      "loss": 0.6349,
+      "step": 4610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8005553728820987,
+      "learning_rate": 8.284285529585526e-06,
+      "loss": 0.6458,
+      "step": 4611
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6713183728884552,
+      "learning_rate": 8.283507628128172e-06,
+      "loss": 0.5637,
+      "step": 4612
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5326842064130279,
+      "learning_rate": 8.282729586902389e-06,
+      "loss": 0.5141,
+      "step": 4613
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7089873735358927,
+      "learning_rate": 8.281951405941298e-06,
+      "loss": 0.5704,
+      "step": 4614
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.748539003985585,
+      "learning_rate": 8.281173085278024e-06,
+      "loss": 0.5747,
+      "step": 4615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.768531314502041,
+      "learning_rate": 8.280394624945697e-06,
+      "loss": 0.5113,
+      "step": 4616
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.596118209725609,
+      "learning_rate": 8.279616024977453e-06,
+      "loss": 0.5245,
+      "step": 4617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8051177654560966,
+      "learning_rate": 8.278837285406436e-06,
+      "loss": 0.5446,
+      "step": 4618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5759237485926437,
+      "learning_rate": 8.278058406265792e-06,
+      "loss": 0.51,
+      "step": 4619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6625567295722475,
+      "learning_rate": 8.277279387588677e-06,
+      "loss": 0.6042,
+      "step": 4620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6024807647703426,
+      "learning_rate": 8.276500229408251e-06,
+      "loss": 0.5292,
+      "step": 4621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7874690897198057,
+      "learning_rate": 8.27572093175768e-06,
+      "loss": 0.6525,
+      "step": 4622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6261995981712538,
+      "learning_rate": 8.27494149467014e-06,
+      "loss": 0.5347,
+      "step": 4623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7515888419419297,
+      "learning_rate": 8.274161918178803e-06,
+      "loss": 0.6693,
+      "step": 4624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6524185402631865,
+      "learning_rate": 8.273382202316855e-06,
+      "loss": 0.5265,
+      "step": 4625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6747795621916263,
+      "learning_rate": 8.272602347117488e-06,
+      "loss": 0.5475,
+      "step": 4626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7020670020957112,
+      "learning_rate": 8.271822352613896e-06,
+      "loss": 0.6218,
+      "step": 4627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5411099911561422,
+      "learning_rate": 8.271042218839284e-06,
+      "loss": 0.5111,
+      "step": 4628
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5493993465983869,
+      "learning_rate": 8.270261945826855e-06,
+      "loss": 0.5359,
+      "step": 4629
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.792438775915959,
+      "learning_rate": 8.269481533609827e-06,
+      "loss": 0.6317,
+      "step": 4630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8024391520749959,
+      "learning_rate": 8.268700982221416e-06,
+      "loss": 0.6572,
+      "step": 4631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9578000117038373,
+      "learning_rate": 8.267920291694851e-06,
+      "loss": 0.5122,
+      "step": 4632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7981641600501669,
+      "learning_rate": 8.267139462063363e-06,
+      "loss": 0.6319,
+      "step": 4633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8111421228066134,
+      "learning_rate": 8.266358493360186e-06,
+      "loss": 0.6145,
+      "step": 4634
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5781219843472877,
+      "learning_rate": 8.265577385618566e-06,
+      "loss": 0.5159,
+      "step": 4635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6945359278874309,
+      "learning_rate": 8.264796138871753e-06,
+      "loss": 0.5012,
+      "step": 4636
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6453162545353781,
+      "learning_rate": 8.264014753153e-06,
+      "loss": 0.5841,
+      "step": 4637
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.666539692458954,
+      "learning_rate": 8.26323322849557e-06,
+      "loss": 0.5674,
+      "step": 4638
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5924871182551501,
+      "learning_rate": 8.262451564932729e-06,
+      "loss": 0.5025,
+      "step": 4639
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6339368540742805,
+      "learning_rate": 8.261669762497752e-06,
+      "loss": 0.5289,
+      "step": 4640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7127930839254408,
+      "learning_rate": 8.260887821223915e-06,
+      "loss": 0.5792,
+      "step": 4641
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7262930078915326,
+      "learning_rate": 8.260105741144502e-06,
+      "loss": 0.5962,
+      "step": 4642
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5935238702267821,
+      "learning_rate": 8.259323522292808e-06,
+      "loss": 0.4399,
+      "step": 4643
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6121339348837413,
+      "learning_rate": 8.258541164702126e-06,
+      "loss": 0.5426,
+      "step": 4644
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.824277049920599,
+      "learning_rate": 8.257758668405758e-06,
+      "loss": 0.6448,
+      "step": 4645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5560645771345775,
+      "learning_rate": 8.256976033437015e-06,
+      "loss": 0.5562,
+      "step": 4646
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6901256974527458,
+      "learning_rate": 8.25619325982921e-06,
+      "loss": 0.5405,
+      "step": 4647
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.779128310073951,
+      "learning_rate": 8.255410347615663e-06,
+      "loss": 0.6679,
+      "step": 4648
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6752717566763254,
+      "learning_rate": 8.2546272968297e-06,
+      "loss": 0.5654,
+      "step": 4649
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8439658033365739,
+      "learning_rate": 8.253844107504653e-06,
+      "loss": 0.6299,
+      "step": 4650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8188284309297219,
+      "learning_rate": 8.253060779673863e-06,
+      "loss": 0.6453,
+      "step": 4651
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7889139822860217,
+      "learning_rate": 8.252277313370668e-06,
+      "loss": 0.6143,
+      "step": 4652
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6781967242411078,
+      "learning_rate": 8.251493708628422e-06,
+      "loss": 0.5723,
+      "step": 4653
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1367495885264813,
+      "learning_rate": 8.250709965480478e-06,
+      "loss": 0.6713,
+      "step": 4654
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5367243516615912,
+      "learning_rate": 8.2499260839602e-06,
+      "loss": 0.537,
+      "step": 4655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7737990284411801,
+      "learning_rate": 8.249142064100952e-06,
+      "loss": 0.5988,
+      "step": 4656
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6965216708231662,
+      "learning_rate": 8.248357905936111e-06,
+      "loss": 0.566,
+      "step": 4657
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7603000484519947,
+      "learning_rate": 8.247573609499054e-06,
+      "loss": 0.5566,
+      "step": 4658
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8556547495436599,
+      "learning_rate": 8.246789174823164e-06,
+      "loss": 0.6139,
+      "step": 4659
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8039671031347132,
+      "learning_rate": 8.246004601941834e-06,
+      "loss": 0.6099,
+      "step": 4660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7857399076760495,
+      "learning_rate": 8.245219890888463e-06,
+      "loss": 0.6252,
+      "step": 4661
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6986999667506865,
+      "learning_rate": 8.244435041696453e-06,
+      "loss": 0.5901,
+      "step": 4662
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6663399209358067,
+      "learning_rate": 8.243650054399209e-06,
+      "loss": 0.5309,
+      "step": 4663
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6288798647661501,
+      "learning_rate": 8.242864929030145e-06,
+      "loss": 0.5573,
+      "step": 4664
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7463076296848864,
+      "learning_rate": 8.242079665622686e-06,
+      "loss": 0.5377,
+      "step": 4665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8739340742472499,
+      "learning_rate": 8.241294264210256e-06,
+      "loss": 0.6134,
+      "step": 4666
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.777090426151362,
+      "learning_rate": 8.240508724826287e-06,
+      "loss": 0.5906,
+      "step": 4667
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.699591831918812,
+      "learning_rate": 8.239723047504216e-06,
+      "loss": 0.5446,
+      "step": 4668
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5941513675925244,
+      "learning_rate": 8.23893723227749e-06,
+      "loss": 0.5582,
+      "step": 4669
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8144842680043904,
+      "learning_rate": 8.238151279179553e-06,
+      "loss": 0.6324,
+      "step": 4670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.785511863474483,
+      "learning_rate": 8.237365188243865e-06,
+      "loss": 0.6472,
+      "step": 4671
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7463472417078749,
+      "learning_rate": 8.236578959503884e-06,
+      "loss": 0.6375,
+      "step": 4672
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6446009794678811,
+      "learning_rate": 8.235792592993082e-06,
+      "loss": 0.5261,
+      "step": 4673
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6573585926396419,
+      "learning_rate": 8.235006088744929e-06,
+      "loss": 0.5246,
+      "step": 4674
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6582561884637631,
+      "learning_rate": 8.234219446792904e-06,
+      "loss": 0.5402,
+      "step": 4675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7633191541450485,
+      "learning_rate": 8.233432667170494e-06,
+      "loss": 0.5676,
+      "step": 4676
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7939556371781813,
+      "learning_rate": 8.232645749911185e-06,
+      "loss": 0.6211,
+      "step": 4677
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.008517296327773,
+      "learning_rate": 8.231858695048479e-06,
+      "loss": 0.6066,
+      "step": 4678
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5938877523332594,
+      "learning_rate": 8.231071502615873e-06,
+      "loss": 0.5422,
+      "step": 4679
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5268837938765629,
+      "learning_rate": 8.230284172646881e-06,
+      "loss": 0.5053,
+      "step": 4680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5833814210144035,
+      "learning_rate": 8.229496705175012e-06,
+      "loss": 0.5064,
+      "step": 4681
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5674897173502002,
+      "learning_rate": 8.228709100233789e-06,
+      "loss": 0.5552,
+      "step": 4682
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5026327757855346,
+      "learning_rate": 8.227921357856738e-06,
+      "loss": 0.463,
+      "step": 4683
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6045577676720874,
+      "learning_rate": 8.22713347807739e-06,
+      "loss": 0.4916,
+      "step": 4684
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6782690097154433,
+      "learning_rate": 8.22634546092928e-06,
+      "loss": 0.5303,
+      "step": 4685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6143614157222634,
+      "learning_rate": 8.225557306445956e-06,
+      "loss": 0.5555,
+      "step": 4686
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.768168823360345,
+      "learning_rate": 8.224769014660964e-06,
+      "loss": 0.5589,
+      "step": 4687
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6340062847522439,
+      "learning_rate": 8.223980585607861e-06,
+      "loss": 0.5835,
+      "step": 4688
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6927697622048208,
+      "learning_rate": 8.223192019320206e-06,
+      "loss": 0.5845,
+      "step": 4689
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5803929956917045,
+      "learning_rate": 8.222403315831565e-06,
+      "loss": 0.5398,
+      "step": 4690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6620576144270319,
+      "learning_rate": 8.221614475175511e-06,
+      "loss": 0.5646,
+      "step": 4691
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6485821328922843,
+      "learning_rate": 8.220825497385628e-06,
+      "loss": 0.5433,
+      "step": 4692
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.758682620589759,
+      "learning_rate": 8.220036382495494e-06,
+      "loss": 0.5635,
+      "step": 4693
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5472629883164765,
+      "learning_rate": 8.219247130538702e-06,
+      "loss": 0.51,
+      "step": 4694
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7145393761055885,
+      "learning_rate": 8.218457741548846e-06,
+      "loss": 0.536,
+      "step": 4695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5911447726164596,
+      "learning_rate": 8.217668215559528e-06,
+      "loss": 0.5062,
+      "step": 4696
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8202246234094444,
+      "learning_rate": 8.216878552604356e-06,
+      "loss": 0.6482,
+      "step": 4697
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7048611594369557,
+      "learning_rate": 8.216088752716945e-06,
+      "loss": 0.5861,
+      "step": 4698
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5855694900257165,
+      "learning_rate": 8.215298815930912e-06,
+      "loss": 0.5338,
+      "step": 4699
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7717944753042869,
+      "learning_rate": 8.214508742279884e-06,
+      "loss": 0.661,
+      "step": 4700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5407674945319926,
+      "learning_rate": 8.21371853179749e-06,
+      "loss": 0.5083,
+      "step": 4701
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7767660691205729,
+      "learning_rate": 8.212928184517368e-06,
+      "loss": 0.5481,
+      "step": 4702
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5305472202996602,
+      "learning_rate": 8.212137700473159e-06,
+      "loss": 0.4873,
+      "step": 4703
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8007867522899832,
+      "learning_rate": 8.211347079698515e-06,
+      "loss": 0.6342,
+      "step": 4704
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5979449171900671,
+      "learning_rate": 8.210556322227087e-06,
+      "loss": 0.5908,
+      "step": 4705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7231031042642851,
+      "learning_rate": 8.209765428092535e-06,
+      "loss": 0.5683,
+      "step": 4706
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7289049404047805,
+      "learning_rate": 8.208974397328527e-06,
+      "loss": 0.6381,
+      "step": 4707
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7236172218250517,
+      "learning_rate": 8.208183229968731e-06,
+      "loss": 0.5567,
+      "step": 4708
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7738967297665412,
+      "learning_rate": 8.20739192604683e-06,
+      "loss": 0.5994,
+      "step": 4709
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7728467453810064,
+      "learning_rate": 8.206600485596501e-06,
+      "loss": 0.6049,
+      "step": 4710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8767782367508747,
+      "learning_rate": 8.205808908651439e-06,
+      "loss": 0.6213,
+      "step": 4711
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7429710637834168,
+      "learning_rate": 8.205017195245333e-06,
+      "loss": 0.6152,
+      "step": 4712
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5897784338241513,
+      "learning_rate": 8.20422534541189e-06,
+      "loss": 0.5291,
+      "step": 4713
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5771387514393338,
+      "learning_rate": 8.203433359184811e-06,
+      "loss": 0.5108,
+      "step": 4714
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8162661306598429,
+      "learning_rate": 8.20264123659781e-06,
+      "loss": 0.5798,
+      "step": 4715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6520075500920509,
+      "learning_rate": 8.201848977684608e-06,
+      "loss": 0.5716,
+      "step": 4716
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9540425134830905,
+      "learning_rate": 8.201056582478926e-06,
+      "loss": 0.6701,
+      "step": 4717
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5588207886438106,
+      "learning_rate": 8.200264051014494e-06,
+      "loss": 0.5205,
+      "step": 4718
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0424848989956124,
+      "learning_rate": 8.199471383325049e-06,
+      "loss": 0.6383,
+      "step": 4719
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6467697823264894,
+      "learning_rate": 8.19867857944433e-06,
+      "loss": 0.5593,
+      "step": 4720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8636816125209708,
+      "learning_rate": 8.197885639406085e-06,
+      "loss": 0.664,
+      "step": 4721
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7121787750414755,
+      "learning_rate": 8.197092563244067e-06,
+      "loss": 0.6414,
+      "step": 4722
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5867722333401326,
+      "learning_rate": 8.196299350992036e-06,
+      "loss": 0.5231,
+      "step": 4723
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8165973481902713,
+      "learning_rate": 8.195506002683756e-06,
+      "loss": 0.6644,
+      "step": 4724
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9932537819337868,
+      "learning_rate": 8.194712518352997e-06,
+      "loss": 0.5714,
+      "step": 4725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7202564353050699,
+      "learning_rate": 8.193918898033533e-06,
+      "loss": 0.6067,
+      "step": 4726
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6208630082852726,
+      "learning_rate": 8.19312514175915e-06,
+      "loss": 0.5114,
+      "step": 4727
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7230895260903912,
+      "learning_rate": 8.192331249563632e-06,
+      "loss": 0.6077,
+      "step": 4728
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5531853516276057,
+      "learning_rate": 8.191537221480775e-06,
+      "loss": 0.4787,
+      "step": 4729
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5858968568748244,
+      "learning_rate": 8.190743057544376e-06,
+      "loss": 0.5306,
+      "step": 4730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.76533443112363,
+      "learning_rate": 8.189948757788242e-06,
+      "loss": 0.6236,
+      "step": 4731
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6381047834579116,
+      "learning_rate": 8.189154322246184e-06,
+      "loss": 0.521,
+      "step": 4732
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8094782169375087,
+      "learning_rate": 8.188359750952015e-06,
+      "loss": 0.622,
+      "step": 4733
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6167793767933879,
+      "learning_rate": 8.187565043939562e-06,
+      "loss": 0.5403,
+      "step": 4734
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8147174576845679,
+      "learning_rate": 8.18677020124265e-06,
+      "loss": 0.6379,
+      "step": 4735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.14250842422233,
+      "learning_rate": 8.185975222895115e-06,
+      "loss": 0.6127,
+      "step": 4736
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6121115758075073,
+      "learning_rate": 8.185180108930795e-06,
+      "loss": 0.5385,
+      "step": 4737
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7378590011021329,
+      "learning_rate": 8.184384859383537e-06,
+      "loss": 0.5876,
+      "step": 4738
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8116117048304031,
+      "learning_rate": 8.18358947428719e-06,
+      "loss": 0.6028,
+      "step": 4739
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7827381595688337,
+      "learning_rate": 8.182793953675613e-06,
+      "loss": 0.5824,
+      "step": 4740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7529424867561774,
+      "learning_rate": 8.181998297582668e-06,
+      "loss": 0.6171,
+      "step": 4741
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8637249480518574,
+      "learning_rate": 8.181202506042224e-06,
+      "loss": 0.6406,
+      "step": 4742
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.84913266113476,
+      "learning_rate": 8.180406579088154e-06,
+      "loss": 0.5968,
+      "step": 4743
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8594606251697315,
+      "learning_rate": 8.179610516754341e-06,
+      "loss": 0.6496,
+      "step": 4744
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5428580741176288,
+      "learning_rate": 8.178814319074669e-06,
+      "loss": 0.5436,
+      "step": 4745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6550157569932077,
+      "learning_rate": 8.178017986083026e-06,
+      "loss": 0.5443,
+      "step": 4746
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7766491928055453,
+      "learning_rate": 8.177221517813317e-06,
+      "loss": 0.5871,
+      "step": 4747
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7229487511566859,
+      "learning_rate": 8.17642491429944e-06,
+      "loss": 0.6646,
+      "step": 4748
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8465769643127085,
+      "learning_rate": 8.175628175575303e-06,
+      "loss": 0.6456,
+      "step": 4749
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9591417564551952,
+      "learning_rate": 8.174831301674823e-06,
+      "loss": 0.6841,
+      "step": 4750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6584652317916735,
+      "learning_rate": 8.17403429263192e-06,
+      "loss": 0.5447,
+      "step": 4751
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6351246382884953,
+      "learning_rate": 8.17323714848052e-06,
+      "loss": 0.4827,
+      "step": 4752
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6794344882886584,
+      "learning_rate": 8.172439869254553e-06,
+      "loss": 0.5947,
+      "step": 4753
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6075077998455788,
+      "learning_rate": 8.171642454987962e-06,
+      "loss": 0.59,
+      "step": 4754
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6062427020311038,
+      "learning_rate": 8.170844905714684e-06,
+      "loss": 0.5528,
+      "step": 4755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7944220427103551,
+      "learning_rate": 8.170047221468673e-06,
+      "loss": 0.6177,
+      "step": 4756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5500138044856991,
+      "learning_rate": 8.16924940228388e-06,
+      "loss": 0.4603,
+      "step": 4757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7967601294356719,
+      "learning_rate": 8.16845144819427e-06,
+      "loss": 0.6302,
+      "step": 4758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5609469815530869,
+      "learning_rate": 8.167653359233803e-06,
+      "loss": 0.5194,
+      "step": 4759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6899560275405495,
+      "learning_rate": 8.166855135436458e-06,
+      "loss": 0.5395,
+      "step": 4760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.640058892806277,
+      "learning_rate": 8.166056776836207e-06,
+      "loss": 0.5185,
+      "step": 4761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7415652042202203,
+      "learning_rate": 8.16525828346704e-06,
+      "loss": 0.6,
+      "step": 4762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7140960492609035,
+      "learning_rate": 8.16445965536294e-06,
+      "loss": 0.5209,
+      "step": 4763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6429796093746752,
+      "learning_rate": 8.163660892557904e-06,
+      "loss": 0.5481,
+      "step": 4764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6797984855087086,
+      "learning_rate": 8.162861995085934e-06,
+      "loss": 0.571,
+      "step": 4765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6390299248708973,
+      "learning_rate": 8.162062962981036e-06,
+      "loss": 0.5633,
+      "step": 4766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5368528206457558,
+      "learning_rate": 8.161263796277223e-06,
+      "loss": 0.4686,
+      "step": 4767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9471618086438857,
+      "learning_rate": 8.160464495008512e-06,
+      "loss": 0.7009,
+      "step": 4768
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6900754661852971,
+      "learning_rate": 8.159665059208926e-06,
+      "loss": 0.563,
+      "step": 4769
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6665293492580482,
+      "learning_rate": 8.158865488912498e-06,
+      "loss": 0.5591,
+      "step": 4770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6719251737210051,
+      "learning_rate": 8.158065784153258e-06,
+      "loss": 0.5602,
+      "step": 4771
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6652603466476381,
+      "learning_rate": 8.15726594496525e-06,
+      "loss": 0.5329,
+      "step": 4772
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6329341996750119,
+      "learning_rate": 8.156465971382518e-06,
+      "loss": 0.5747,
+      "step": 4773
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7178080928686216,
+      "learning_rate": 8.155665863439118e-06,
+      "loss": 0.5932,
+      "step": 4774
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5875864628155599,
+      "learning_rate": 8.154865621169106e-06,
+      "loss": 0.5702,
+      "step": 4775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.703005532628864,
+      "learning_rate": 8.154065244606547e-06,
+      "loss": 0.5411,
+      "step": 4776
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7231004309489556,
+      "learning_rate": 8.153264733785506e-06,
+      "loss": 0.6325,
+      "step": 4777
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5795419795889204,
+      "learning_rate": 8.152464088740066e-06,
+      "loss": 0.5075,
+      "step": 4778
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.975411624673703,
+      "learning_rate": 8.151663309504301e-06,
+      "loss": 0.6291,
+      "step": 4779
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6978743100613516,
+      "learning_rate": 8.1508623961123e-06,
+      "loss": 0.5392,
+      "step": 4780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7707502929035336,
+      "learning_rate": 8.150061348598158e-06,
+      "loss": 0.6428,
+      "step": 4781
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7465758153972254,
+      "learning_rate": 8.14926016699597e-06,
+      "loss": 0.55,
+      "step": 4782
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6830603107167149,
+      "learning_rate": 8.148458851339837e-06,
+      "loss": 0.5929,
+      "step": 4783
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6912109698268275,
+      "learning_rate": 8.147657401663875e-06,
+      "loss": 0.5323,
+      "step": 4784
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6295671870593303,
+      "learning_rate": 8.146855818002194e-06,
+      "loss": 0.547,
+      "step": 4785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5388794576131567,
+      "learning_rate": 8.146054100388917e-06,
+      "loss": 0.497,
+      "step": 4786
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.561849712112106,
+      "learning_rate": 8.145252248858172e-06,
+      "loss": 0.4933,
+      "step": 4787
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6238080545111447,
+      "learning_rate": 8.144450263444089e-06,
+      "loss": 0.4867,
+      "step": 4788
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6100209454034814,
+      "learning_rate": 8.143648144180804e-06,
+      "loss": 0.5575,
+      "step": 4789
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6869051626211036,
+      "learning_rate": 8.142845891102466e-06,
+      "loss": 0.5246,
+      "step": 4790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7185336470393379,
+      "learning_rate": 8.14204350424322e-06,
+      "loss": 0.5567,
+      "step": 4791
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7722515859289103,
+      "learning_rate": 8.141240983637224e-06,
+      "loss": 0.5818,
+      "step": 4792
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7579152851004997,
+      "learning_rate": 8.140438329318636e-06,
+      "loss": 0.6249,
+      "step": 4793
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.718007431908234,
+      "learning_rate": 8.139635541321624e-06,
+      "loss": 0.5953,
+      "step": 4794
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6915713114399602,
+      "learning_rate": 8.138832619680361e-06,
+      "loss": 0.5483,
+      "step": 4795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.826928163300088,
+      "learning_rate": 8.138029564429022e-06,
+      "loss": 0.6276,
+      "step": 4796
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6821590812335693,
+      "learning_rate": 8.137226375601793e-06,
+      "loss": 0.5521,
+      "step": 4797
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7539433385643475,
+      "learning_rate": 8.136423053232863e-06,
+      "loss": 0.6502,
+      "step": 4798
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8280041133982782,
+      "learning_rate": 8.135619597356426e-06,
+      "loss": 0.5772,
+      "step": 4799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7759999067614988,
+      "learning_rate": 8.134816008006681e-06,
+      "loss": 0.6089,
+      "step": 4800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5967147981107758,
+      "learning_rate": 8.134012285217838e-06,
+      "loss": 0.5352,
+      "step": 4801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.605268819827656,
+      "learning_rate": 8.133208429024106e-06,
+      "loss": 0.5731,
+      "step": 4802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7784713763142074,
+      "learning_rate": 8.132404439459702e-06,
+      "loss": 0.5946,
+      "step": 4803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6422953028824958,
+      "learning_rate": 8.131600316558852e-06,
+      "loss": 0.5425,
+      "step": 4804
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9488262430014444,
+      "learning_rate": 8.130796060355784e-06,
+      "loss": 0.5849,
+      "step": 4805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.760895490255571,
+      "learning_rate": 8.129991670884733e-06,
+      "loss": 0.6297,
+      "step": 4806
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.797257959647494,
+      "learning_rate": 8.129187148179938e-06,
+      "loss": 0.5763,
+      "step": 4807
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6082905580279685,
+      "learning_rate": 8.128382492275644e-06,
+      "loss": 0.5502,
+      "step": 4808
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6848614964915353,
+      "learning_rate": 8.127577703206106e-06,
+      "loss": 0.6208,
+      "step": 4809
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9584029166809982,
+      "learning_rate": 8.12677278100558e-06,
+      "loss": 0.6336,
+      "step": 4810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7183629341738105,
+      "learning_rate": 8.125967725708328e-06,
+      "loss": 0.5912,
+      "step": 4811
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6785159993765555,
+      "learning_rate": 8.125162537348619e-06,
+      "loss": 0.5499,
+      "step": 4812
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8716339175559804,
+      "learning_rate": 8.124357215960727e-06,
+      "loss": 0.6399,
+      "step": 4813
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6433837587231448,
+      "learning_rate": 8.123551761578931e-06,
+      "loss": 0.5176,
+      "step": 4814
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6443099290245619,
+      "learning_rate": 8.122746174237518e-06,
+      "loss": 0.5517,
+      "step": 4815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8781899269543718,
+      "learning_rate": 8.121940453970783e-06,
+      "loss": 0.6455,
+      "step": 4816
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6489200588777149,
+      "learning_rate": 8.121134600813015e-06,
+      "loss": 0.5344,
+      "step": 4817
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6396351274396583,
+      "learning_rate": 8.120328614798523e-06,
+      "loss": 0.5311,
+      "step": 4818
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7540203510791869,
+      "learning_rate": 8.119522495961612e-06,
+      "loss": 0.6351,
+      "step": 4819
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6818248485252133,
+      "learning_rate": 8.118716244336597e-06,
+      "loss": 0.5111,
+      "step": 4820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6778605624073916,
+      "learning_rate": 8.117909859957798e-06,
+      "loss": 0.5092,
+      "step": 4821
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6924205642203811,
+      "learning_rate": 8.11710334285954e-06,
+      "loss": 0.5422,
+      "step": 4822
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5901394765861796,
+      "learning_rate": 8.116296693076154e-06,
+      "loss": 0.5939,
+      "step": 4823
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8388903654904268,
+      "learning_rate": 8.115489910641974e-06,
+      "loss": 0.6289,
+      "step": 4824
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6627781865118006,
+      "learning_rate": 8.114682995591345e-06,
+      "loss": 0.5172,
+      "step": 4825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5900976791764801,
+      "learning_rate": 8.113875947958616e-06,
+      "loss": 0.5203,
+      "step": 4826
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6738291442510475,
+      "learning_rate": 8.113068767778134e-06,
+      "loss": 0.5559,
+      "step": 4827
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.566831010030039,
+      "learning_rate": 8.112261455084265e-06,
+      "loss": 0.4868,
+      "step": 4828
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6321945141159364,
+      "learning_rate": 8.111454009911372e-06,
+      "loss": 0.563,
+      "step": 4829
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.635862688232433,
+      "learning_rate": 8.110646432293824e-06,
+      "loss": 0.5428,
+      "step": 4830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7577421420038001,
+      "learning_rate": 8.109838722265997e-06,
+      "loss": 0.6351,
+      "step": 4831
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7286074056901038,
+      "learning_rate": 8.109030879862274e-06,
+      "loss": 0.5856,
+      "step": 4832
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.633054677055702,
+      "learning_rate": 8.108222905117042e-06,
+      "loss": 0.5816,
+      "step": 4833
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6770640871206239,
+      "learning_rate": 8.107414798064691e-06,
+      "loss": 0.5155,
+      "step": 4834
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7535503897137129,
+      "learning_rate": 8.106606558739625e-06,
+      "loss": 0.6265,
+      "step": 4835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7024832634606328,
+      "learning_rate": 8.105798187176244e-06,
+      "loss": 0.553,
+      "step": 4836
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8754086666788395,
+      "learning_rate": 8.104989683408958e-06,
+      "loss": 0.633,
+      "step": 4837
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7669117862269665,
+      "learning_rate": 8.104181047472184e-06,
+      "loss": 0.6301,
+      "step": 4838
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9213269387667131,
+      "learning_rate": 8.103372279400341e-06,
+      "loss": 0.6266,
+      "step": 4839
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6657568430405255,
+      "learning_rate": 8.10256337922786e-06,
+      "loss": 0.5311,
+      "step": 4840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7967500853464504,
+      "learning_rate": 8.101754346989168e-06,
+      "loss": 0.6059,
+      "step": 4841
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.597754659123826,
+      "learning_rate": 8.100945182718705e-06,
+      "loss": 0.5488,
+      "step": 4842
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6263617389601803,
+      "learning_rate": 8.100135886450916e-06,
+      "loss": 0.651,
+      "step": 4843
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7260252159977802,
+      "learning_rate": 8.099326458220249e-06,
+      "loss": 0.5937,
+      "step": 4844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2379358999520154,
+      "learning_rate": 8.098516898061159e-06,
+      "loss": 0.6083,
+      "step": 4845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5925252654182341,
+      "learning_rate": 8.097707206008102e-06,
+      "loss": 0.5397,
+      "step": 4846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6085251850615595,
+      "learning_rate": 8.096897382095553e-06,
+      "loss": 0.5117,
+      "step": 4847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7328260826979824,
+      "learning_rate": 8.096087426357978e-06,
+      "loss": 0.6174,
+      "step": 4848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6541158758973585,
+      "learning_rate": 8.095277338829853e-06,
+      "loss": 0.5069,
+      "step": 4849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5777393348145821,
+      "learning_rate": 8.094467119545663e-06,
+      "loss": 0.5154,
+      "step": 4850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7090133563575965,
+      "learning_rate": 8.0936567685399e-06,
+      "loss": 0.5341,
+      "step": 4851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7579490527579456,
+      "learning_rate": 8.092846285847049e-06,
+      "loss": 0.6279,
+      "step": 4852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6256002701802358,
+      "learning_rate": 8.092035671501617e-06,
+      "loss": 0.5118,
+      "step": 4853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6212915649896393,
+      "learning_rate": 8.091224925538108e-06,
+      "loss": 0.5432,
+      "step": 4854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5667830327736616,
+      "learning_rate": 8.090414047991031e-06,
+      "loss": 0.4996,
+      "step": 4855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7373447686309222,
+      "learning_rate": 8.089603038894904e-06,
+      "loss": 0.5686,
+      "step": 4856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9525439248404798,
+      "learning_rate": 8.08879189828425e-06,
+      "loss": 0.5795,
+      "step": 4857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7785048846145941,
+      "learning_rate": 8.087980626193592e-06,
+      "loss": 0.6069,
+      "step": 4858
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6276731033295745,
+      "learning_rate": 8.08716922265747e-06,
+      "loss": 0.6157,
+      "step": 4859
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6201341741204739,
+      "learning_rate": 8.086357687710417e-06,
+      "loss": 0.5355,
+      "step": 4860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8459649364744836,
+      "learning_rate": 8.08554602138698e-06,
+      "loss": 0.6384,
+      "step": 4861
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7000963566777608,
+      "learning_rate": 8.08473422372171e-06,
+      "loss": 0.6873,
+      "step": 4862
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6653633755268963,
+      "learning_rate": 8.08392229474916e-06,
+      "loss": 0.5585,
+      "step": 4863
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6239248567406012,
+      "learning_rate": 8.083110234503895e-06,
+      "loss": 0.5455,
+      "step": 4864
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7981524243795168,
+      "learning_rate": 8.082298043020478e-06,
+      "loss": 0.6161,
+      "step": 4865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7162597540883014,
+      "learning_rate": 8.081485720333482e-06,
+      "loss": 0.5786,
+      "step": 4866
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6799110198446136,
+      "learning_rate": 8.080673266477488e-06,
+      "loss": 0.5397,
+      "step": 4867
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7345853282228795,
+      "learning_rate": 8.079860681487076e-06,
+      "loss": 0.59,
+      "step": 4868
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6141263424194727,
+      "learning_rate": 8.079047965396839e-06,
+      "loss": 0.5624,
+      "step": 4869
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5948313931876724,
+      "learning_rate": 8.078235118241365e-06,
+      "loss": 0.5591,
+      "step": 4870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8005217658260888,
+      "learning_rate": 8.077422140055261e-06,
+      "loss": 0.6141,
+      "step": 4871
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5711097481074564,
+      "learning_rate": 8.07660903087313e-06,
+      "loss": 0.5394,
+      "step": 4872
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7340286153341028,
+      "learning_rate": 8.075795790729586e-06,
+      "loss": 0.6156,
+      "step": 4873
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7820995091205901,
+      "learning_rate": 8.074982419659244e-06,
+      "loss": 0.5836,
+      "step": 4874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6910957990008223,
+      "learning_rate": 8.074168917696724e-06,
+      "loss": 0.5851,
+      "step": 4875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8480003491441266,
+      "learning_rate": 8.073355284876658e-06,
+      "loss": 0.5989,
+      "step": 4876
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8737900651445702,
+      "learning_rate": 8.072541521233679e-06,
+      "loss": 0.6439,
+      "step": 4877
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8007813808033702,
+      "learning_rate": 8.071727626802424e-06,
+      "loss": 0.6284,
+      "step": 4878
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6936560572133352,
+      "learning_rate": 8.070913601617543e-06,
+      "loss": 0.5069,
+      "step": 4879
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7981425101843691,
+      "learning_rate": 8.07009944571368e-06,
+      "loss": 0.6347,
+      "step": 4880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5779411834405241,
+      "learning_rate": 8.069285159125496e-06,
+      "loss": 0.5422,
+      "step": 4881
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8190092659255641,
+      "learning_rate": 8.068470741887651e-06,
+      "loss": 0.6454,
+      "step": 4882
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.804109598707946,
+      "learning_rate": 8.067656194034811e-06,
+      "loss": 0.5844,
+      "step": 4883
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5820104262480262,
+      "learning_rate": 8.06684151560165e-06,
+      "loss": 0.5529,
+      "step": 4884
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6473168890837421,
+      "learning_rate": 8.066026706622847e-06,
+      "loss": 0.5917,
+      "step": 4885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7896002476119041,
+      "learning_rate": 8.065211767133084e-06,
+      "loss": 0.6267,
+      "step": 4886
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8729663423341352,
+      "learning_rate": 8.06439669716705e-06,
+      "loss": 0.683,
+      "step": 4887
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8224710341287688,
+      "learning_rate": 8.063581496759443e-06,
+      "loss": 0.6236,
+      "step": 4888
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6662888533544024,
+      "learning_rate": 8.06276616594496e-06,
+      "loss": 0.5263,
+      "step": 4889
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9789375430210324,
+      "learning_rate": 8.061950704758307e-06,
+      "loss": 0.6493,
+      "step": 4890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8049248495167562,
+      "learning_rate": 8.061135113234199e-06,
+      "loss": 0.5939,
+      "step": 4891
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.683245514292312,
+      "learning_rate": 8.060319391407352e-06,
+      "loss": 0.5611,
+      "step": 4892
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5641700951459367,
+      "learning_rate": 8.059503539312487e-06,
+      "loss": 0.5655,
+      "step": 4893
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5811884318442994,
+      "learning_rate": 8.058687556984333e-06,
+      "loss": 0.5113,
+      "step": 4894
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5629376793973534,
+      "learning_rate": 8.057871444457624e-06,
+      "loss": 0.4969,
+      "step": 4895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6015200960107091,
+      "learning_rate": 8.057055201767098e-06,
+      "loss": 0.5512,
+      "step": 4896
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6332036574827241,
+      "learning_rate": 8.056238828947503e-06,
+      "loss": 0.5584,
+      "step": 4897
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5763057480957622,
+      "learning_rate": 8.055422326033586e-06,
+      "loss": 0.5248,
+      "step": 4898
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8874959396134705,
+      "learning_rate": 8.054605693060106e-06,
+      "loss": 0.6002,
+      "step": 4899
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6794838650061711,
+      "learning_rate": 8.053788930061822e-06,
+      "loss": 0.5925,
+      "step": 4900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8484945740126925,
+      "learning_rate": 8.052972037073501e-06,
+      "loss": 0.6414,
+      "step": 4901
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.612939219459935,
+      "learning_rate": 8.052155014129916e-06,
+      "loss": 0.5461,
+      "step": 4902
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7431409948859721,
+      "learning_rate": 8.051337861265847e-06,
+      "loss": 0.5727,
+      "step": 4903
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5868836082094497,
+      "learning_rate": 8.050520578516074e-06,
+      "loss": 0.5204,
+      "step": 4904
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6702798902963087,
+      "learning_rate": 8.04970316591539e-06,
+      "loss": 0.5689,
+      "step": 4905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7830835168618177,
+      "learning_rate": 8.048885623498587e-06,
+      "loss": 0.6054,
+      "step": 4906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7448351170870566,
+      "learning_rate": 8.048067951300465e-06,
+      "loss": 0.6233,
+      "step": 4907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6486517744835938,
+      "learning_rate": 8.04725014935583e-06,
+      "loss": 0.554,
+      "step": 4908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6722504772304511,
+      "learning_rate": 8.046432217699496e-06,
+      "loss": 0.5193,
+      "step": 4909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8182603359914264,
+      "learning_rate": 8.045614156366276e-06,
+      "loss": 0.6271,
+      "step": 4910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6446605809609107,
+      "learning_rate": 8.044795965390995e-06,
+      "loss": 0.5534,
+      "step": 4911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7035520761978654,
+      "learning_rate": 8.043977644808478e-06,
+      "loss": 0.5389,
+      "step": 4912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.827594038147539,
+      "learning_rate": 8.04315919465356e-06,
+      "loss": 0.6375,
+      "step": 4913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7374695926988003,
+      "learning_rate": 8.04234061496108e-06,
+      "loss": 0.6404,
+      "step": 4914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5693042303910528,
+      "learning_rate": 8.041521905765883e-06,
+      "loss": 0.5619,
+      "step": 4915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5875885745632737,
+      "learning_rate": 8.040703067102816e-06,
+      "loss": 0.5437,
+      "step": 4916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7185821643081933,
+      "learning_rate": 8.039884099006739e-06,
+      "loss": 0.5417,
+      "step": 4917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6892261854689584,
+      "learning_rate": 8.039065001512508e-06,
+      "loss": 0.5838,
+      "step": 4918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6964076157822582,
+      "learning_rate": 8.03824577465499e-06,
+      "loss": 0.5413,
+      "step": 4919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7527568758841825,
+      "learning_rate": 8.037426418469058e-06,
+      "loss": 0.6089,
+      "step": 4920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7509769115590305,
+      "learning_rate": 8.036606932989592e-06,
+      "loss": 0.5368,
+      "step": 4921
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5815321045061785,
+      "learning_rate": 8.03578731825147e-06,
+      "loss": 0.4913,
+      "step": 4922
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7236216992126074,
+      "learning_rate": 8.034967574289587e-06,
+      "loss": 0.6146,
+      "step": 4923
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7644428638996943,
+      "learning_rate": 8.03414770113883e-06,
+      "loss": 0.528,
+      "step": 4924
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6993079172656466,
+      "learning_rate": 8.033327698834101e-06,
+      "loss": 0.5754,
+      "step": 4925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7313450764515096,
+      "learning_rate": 8.032507567410304e-06,
+      "loss": 0.6255,
+      "step": 4926
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.761825300077517,
+      "learning_rate": 8.031687306902352e-06,
+      "loss": 0.6288,
+      "step": 4927
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6141230843182297,
+      "learning_rate": 8.030866917345159e-06,
+      "loss": 0.5166,
+      "step": 4928
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6644433915350286,
+      "learning_rate": 8.030046398773647e-06,
+      "loss": 0.5611,
+      "step": 4929
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.573637497473543,
+      "learning_rate": 8.029225751222743e-06,
+      "loss": 0.5396,
+      "step": 4930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6878188589679671,
+      "learning_rate": 8.028404974727377e-06,
+      "loss": 0.5248,
+      "step": 4931
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8069764538185773,
+      "learning_rate": 8.02758406932249e-06,
+      "loss": 0.6407,
+      "step": 4932
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6955927656844051,
+      "learning_rate": 8.026763035043021e-06,
+      "loss": 0.5512,
+      "step": 4933
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7575647114305473,
+      "learning_rate": 8.025941871923926e-06,
+      "loss": 0.5891,
+      "step": 4934
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6995656652211012,
+      "learning_rate": 8.025120580000153e-06,
+      "loss": 0.5557,
+      "step": 4935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7235148605292684,
+      "learning_rate": 8.024299159306663e-06,
+      "loss": 0.6974,
+      "step": 4936
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6035609518052024,
+      "learning_rate": 8.023477609878422e-06,
+      "loss": 0.5044,
+      "step": 4937
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5910437522974566,
+      "learning_rate": 8.022655931750403e-06,
+      "loss": 0.5683,
+      "step": 4938
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6474405146828806,
+      "learning_rate": 8.021834124957578e-06,
+      "loss": 0.5213,
+      "step": 4939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8086742935312414,
+      "learning_rate": 8.021012189534931e-06,
+      "loss": 0.6436,
+      "step": 4940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.616790651415157,
+      "learning_rate": 8.020190125517451e-06,
+      "loss": 0.5632,
+      "step": 4941
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9826153405074254,
+      "learning_rate": 8.019367932940126e-06,
+      "loss": 0.6423,
+      "step": 4942
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7436441861685466,
+      "learning_rate": 8.018545611837959e-06,
+      "loss": 0.6507,
+      "step": 4943
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4778030251635295,
+      "learning_rate": 8.017723162245948e-06,
+      "loss": 0.6036,
+      "step": 4944
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6638589947665885,
+      "learning_rate": 8.016900584199106e-06,
+      "loss": 0.5245,
+      "step": 4945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.580197713064733,
+      "learning_rate": 8.016077877732448e-06,
+      "loss": 0.5446,
+      "step": 4946
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6682971371288308,
+      "learning_rate": 8.015255042880993e-06,
+      "loss": 0.5816,
+      "step": 4947
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7183818744936854,
+      "learning_rate": 8.014432079679766e-06,
+      "loss": 0.5814,
+      "step": 4948
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.685527682626321,
+      "learning_rate": 8.013608988163797e-06,
+      "loss": 0.6078,
+      "step": 4949
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.59734172242765,
+      "learning_rate": 8.012785768368125e-06,
+      "loss": 0.5283,
+      "step": 4950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6447979255046242,
+      "learning_rate": 8.011962420327788e-06,
+      "loss": 0.604,
+      "step": 4951
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.809380976290288,
+      "learning_rate": 8.011138944077838e-06,
+      "loss": 0.6925,
+      "step": 4952
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7869086565648559,
+      "learning_rate": 8.010315339653325e-06,
+      "loss": 0.6037,
+      "step": 4953
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7987426487144512,
+      "learning_rate": 8.009491607089305e-06,
+      "loss": 0.6005,
+      "step": 4954
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7516274719269177,
+      "learning_rate": 8.008667746420849e-06,
+      "loss": 0.6318,
+      "step": 4955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7740498530156996,
+      "learning_rate": 8.007843757683019e-06,
+      "loss": 0.598,
+      "step": 4956
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6904989824350313,
+      "learning_rate": 8.00701964091089e-06,
+      "loss": 0.5592,
+      "step": 4957
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6199203109786972,
+      "learning_rate": 8.006195396139545e-06,
+      "loss": 0.5515,
+      "step": 4958
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8311154963488638,
+      "learning_rate": 8.00537102340407e-06,
+      "loss": 0.6265,
+      "step": 4959
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.673649226032902,
+      "learning_rate": 8.004546522739553e-06,
+      "loss": 0.5239,
+      "step": 4960
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9741430450093904,
+      "learning_rate": 8.003721894181092e-06,
+      "loss": 0.596,
+      "step": 4961
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7731955285755833,
+      "learning_rate": 8.00289713776379e-06,
+      "loss": 0.6207,
+      "step": 4962
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8706968197408063,
+      "learning_rate": 8.00207225352275e-06,
+      "loss": 0.6353,
+      "step": 4963
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8210021030467289,
+      "learning_rate": 8.001247241493089e-06,
+      "loss": 0.6399,
+      "step": 4964
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6537251604318643,
+      "learning_rate": 8.000422101709923e-06,
+      "loss": 0.5753,
+      "step": 4965
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9390043673291344,
+      "learning_rate": 7.999596834208377e-06,
+      "loss": 0.6659,
+      "step": 4966
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.963241817320959,
+      "learning_rate": 7.998771439023578e-06,
+      "loss": 0.6748,
+      "step": 4967
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.789801755202053,
+      "learning_rate": 7.997945916190661e-06,
+      "loss": 0.6539,
+      "step": 4968
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7281406478570932,
+      "learning_rate": 7.99712026574477e-06,
+      "loss": 0.5985,
+      "step": 4969
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7085923490566987,
+      "learning_rate": 7.996294487721041e-06,
+      "loss": 0.5913,
+      "step": 4970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6460091542718994,
+      "learning_rate": 7.995468582154634e-06,
+      "loss": 0.4897,
+      "step": 4971
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.624189856220877,
+      "learning_rate": 7.994642549080702e-06,
+      "loss": 0.5663,
+      "step": 4972
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7484498850374535,
+      "learning_rate": 7.993816388534404e-06,
+      "loss": 0.5894,
+      "step": 4973
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.81555407216843,
+      "learning_rate": 7.99299010055091e-06,
+      "loss": 0.5471,
+      "step": 4974
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.574974637473329,
+      "learning_rate": 7.992163685165393e-06,
+      "loss": 0.5283,
+      "step": 4975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6329118697095062,
+      "learning_rate": 7.991337142413029e-06,
+      "loss": 0.5339,
+      "step": 4976
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6164356332347005,
+      "learning_rate": 7.990510472329e-06,
+      "loss": 0.5469,
+      "step": 4977
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6144438291787048,
+      "learning_rate": 7.989683674948498e-06,
+      "loss": 0.4932,
+      "step": 4978
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8879003383011629,
+      "learning_rate": 7.988856750306716e-06,
+      "loss": 0.6502,
+      "step": 4979
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7075341439946503,
+      "learning_rate": 7.988029698438853e-06,
+      "loss": 0.5467,
+      "step": 4980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.651597186699634,
+      "learning_rate": 7.987202519380114e-06,
+      "loss": 0.5461,
+      "step": 4981
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8258889716800095,
+      "learning_rate": 7.98637521316571e-06,
+      "loss": 0.6369,
+      "step": 4982
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6822670076700751,
+      "learning_rate": 7.985547779830856e-06,
+      "loss": 0.5853,
+      "step": 4983
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6908581928240837,
+      "learning_rate": 7.984720219410773e-06,
+      "loss": 0.5748,
+      "step": 4984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9160495837039937,
+      "learning_rate": 7.983892531940687e-06,
+      "loss": 0.6137,
+      "step": 4985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5555883207325232,
+      "learning_rate": 7.983064717455832e-06,
+      "loss": 0.4661,
+      "step": 4986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6066013306324208,
+      "learning_rate": 7.982236775991445e-06,
+      "loss": 0.5452,
+      "step": 4987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6284802021881586,
+      "learning_rate": 7.981408707582769e-06,
+      "loss": 0.5563,
+      "step": 4988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6886870793431915,
+      "learning_rate": 7.98058051226505e-06,
+      "loss": 0.6052,
+      "step": 4989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6990423918403783,
+      "learning_rate": 7.979752190073543e-06,
+      "loss": 0.551,
+      "step": 4990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7674093938950507,
+      "learning_rate": 7.978923741043508e-06,
+      "loss": 0.5451,
+      "step": 4991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.199156064081174,
+      "learning_rate": 7.978095165210209e-06,
+      "loss": 0.6322,
+      "step": 4992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6908099319990278,
+      "learning_rate": 7.977266462608915e-06,
+      "loss": 0.5999,
+      "step": 4993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6431466981335197,
+      "learning_rate": 7.976437633274901e-06,
+      "loss": 0.5218,
+      "step": 4994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6817665191447893,
+      "learning_rate": 7.975608677243449e-06,
+      "loss": 0.5663,
+      "step": 4995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6648658984445288,
+      "learning_rate": 7.974779594549844e-06,
+      "loss": 0.5393,
+      "step": 4996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6637238420430308,
+      "learning_rate": 7.973950385229378e-06,
+      "loss": 0.5556,
+      "step": 4997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8176644553156135,
+      "learning_rate": 7.973121049317349e-06,
+      "loss": 0.6779,
+      "step": 4998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6876679393660092,
+      "learning_rate": 7.972291586849054e-06,
+      "loss": 0.5757,
+      "step": 4999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6717052628132845,
+      "learning_rate": 7.971461997859808e-06,
+      "loss": 0.5094,
+      "step": 5000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7194117786502368,
+      "learning_rate": 7.970632282384918e-06,
+      "loss": 0.5393,
+      "step": 5001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0149557103812121,
+      "learning_rate": 7.969802440459704e-06,
+      "loss": 0.599,
+      "step": 5002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6942619217568177,
+      "learning_rate": 7.968972472119491e-06,
+      "loss": 0.5802,
+      "step": 5003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5765078417775431,
+      "learning_rate": 7.968142377399608e-06,
+      "loss": 0.4968,
+      "step": 5004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.811633019565186,
+      "learning_rate": 7.967312156335389e-06,
+      "loss": 0.6265,
+      "step": 5005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8410833284701286,
+      "learning_rate": 7.966481808962174e-06,
+      "loss": 0.6364,
+      "step": 5006
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7550709211650153,
+      "learning_rate": 7.965651335315305e-06,
+      "loss": 0.5816,
+      "step": 5007
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6682562657334422,
+      "learning_rate": 7.96482073543014e-06,
+      "loss": 0.5109,
+      "step": 5008
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7032653731154704,
+      "learning_rate": 7.963990009342026e-06,
+      "loss": 0.5968,
+      "step": 5009
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6804902791575567,
+      "learning_rate": 7.96315915708633e-06,
+      "loss": 0.5686,
+      "step": 5010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7233130333012316,
+      "learning_rate": 7.96232817869842e-06,
+      "loss": 0.6483,
+      "step": 5011
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5943661914938109,
+      "learning_rate": 7.961497074213664e-06,
+      "loss": 0.5234,
+      "step": 5012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7687568441710754,
+      "learning_rate": 7.96066584366744e-06,
+      "loss": 0.6158,
+      "step": 5013
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7907268912226469,
+      "learning_rate": 7.959834487095135e-06,
+      "loss": 0.5825,
+      "step": 5014
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8253108962829002,
+      "learning_rate": 7.959003004532132e-06,
+      "loss": 0.6264,
+      "step": 5015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8314557847227163,
+      "learning_rate": 7.958171396013827e-06,
+      "loss": 0.6159,
+      "step": 5016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6070493284253823,
+      "learning_rate": 7.957339661575618e-06,
+      "loss": 0.5816,
+      "step": 5017
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8312694095835157,
+      "learning_rate": 7.956507801252912e-06,
+      "loss": 0.6823,
+      "step": 5018
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7910304487107016,
+      "learning_rate": 7.955675815081114e-06,
+      "loss": 0.5786,
+      "step": 5019
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6418403797533321,
+      "learning_rate": 7.954843703095644e-06,
+      "loss": 0.571,
+      "step": 5020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9557562828315895,
+      "learning_rate": 7.954011465331917e-06,
+      "loss": 0.5964,
+      "step": 5021
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7629141335309464,
+      "learning_rate": 7.953179101825364e-06,
+      "loss": 0.5923,
+      "step": 5022
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5695493872500682,
+      "learning_rate": 7.952346612611413e-06,
+      "loss": 0.5009,
+      "step": 5023
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5967780145596528,
+      "learning_rate": 7.951513997725501e-06,
+      "loss": 0.5137,
+      "step": 5024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6046722760236224,
+      "learning_rate": 7.95068125720307e-06,
+      "loss": 0.5109,
+      "step": 5025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9257658633257344,
+      "learning_rate": 7.949848391079566e-06,
+      "loss": 0.6057,
+      "step": 5026
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7817539946449922,
+      "learning_rate": 7.949015399390443e-06,
+      "loss": 0.637,
+      "step": 5027
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5977205782331978,
+      "learning_rate": 7.94818228217116e-06,
+      "loss": 0.511,
+      "step": 5028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.845289093357311,
+      "learning_rate": 7.947349039457175e-06,
+      "loss": 0.6272,
+      "step": 5029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6242821350284501,
+      "learning_rate": 7.946515671283962e-06,
+      "loss": 0.5621,
+      "step": 5030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.740643890292027,
+      "learning_rate": 7.945682177686992e-06,
+      "loss": 0.5391,
+      "step": 5031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6881059538004348,
+      "learning_rate": 7.944848558701743e-06,
+      "loss": 0.6137,
+      "step": 5032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5642095014106421,
+      "learning_rate": 7.944014814363703e-06,
+      "loss": 0.514,
+      "step": 5033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6670245561237913,
+      "learning_rate": 7.943180944708361e-06,
+      "loss": 0.6139,
+      "step": 5034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7789270633283152,
+      "learning_rate": 7.942346949771211e-06,
+      "loss": 0.5987,
+      "step": 5035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.572422754598143,
+      "learning_rate": 7.941512829587753e-06,
+      "loss": 0.5582,
+      "step": 5036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5562236654922782,
+      "learning_rate": 7.940678584193492e-06,
+      "loss": 0.5599,
+      "step": 5037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6948678762479501,
+      "learning_rate": 7.939844213623942e-06,
+      "loss": 0.5732,
+      "step": 5038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7105738402823709,
+      "learning_rate": 7.939009717914619e-06,
+      "loss": 0.6254,
+      "step": 5039
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6884141398825876,
+      "learning_rate": 7.938175097101043e-06,
+      "loss": 0.5349,
+      "step": 5040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7075112009203091,
+      "learning_rate": 7.937340351218743e-06,
+      "loss": 0.5585,
+      "step": 5041
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5841872756180618,
+      "learning_rate": 7.936505480303251e-06,
+      "loss": 0.5211,
+      "step": 5042
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7582660928114365,
+      "learning_rate": 7.935670484390103e-06,
+      "loss": 0.5876,
+      "step": 5043
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6394882090662029,
+      "learning_rate": 7.934835363514843e-06,
+      "loss": 0.564,
+      "step": 5044
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5738471337815415,
+      "learning_rate": 7.934000117713021e-06,
+      "loss": 0.5109,
+      "step": 5045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6522987781055405,
+      "learning_rate": 7.933164747020191e-06,
+      "loss": 0.5848,
+      "step": 5046
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7265418727142777,
+      "learning_rate": 7.932329251471911e-06,
+      "loss": 0.509,
+      "step": 5047
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7495131456050885,
+      "learning_rate": 7.931493631103743e-06,
+      "loss": 0.6102,
+      "step": 5048
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7328420990163743,
+      "learning_rate": 7.930657885951261e-06,
+      "loss": 0.5479,
+      "step": 5049
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.620422845891505,
+      "learning_rate": 7.929822016050034e-06,
+      "loss": 0.5732,
+      "step": 5050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7642114420941445,
+      "learning_rate": 7.92898602143565e-06,
+      "loss": 0.6453,
+      "step": 5051
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7367152211775877,
+      "learning_rate": 7.92814990214369e-06,
+      "loss": 0.5673,
+      "step": 5052
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53367041818719,
+      "learning_rate": 7.927313658209744e-06,
+      "loss": 0.5007,
+      "step": 5053
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7977123909340416,
+      "learning_rate": 7.926477289669411e-06,
+      "loss": 0.6518,
+      "step": 5054
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8386176978568222,
+      "learning_rate": 7.925640796558291e-06,
+      "loss": 0.6049,
+      "step": 5055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7815541597991627,
+      "learning_rate": 7.924804178911993e-06,
+      "loss": 0.579,
+      "step": 5056
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7325344772733176,
+      "learning_rate": 7.923967436766126e-06,
+      "loss": 0.55,
+      "step": 5057
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7146358864359678,
+      "learning_rate": 7.92313057015631e-06,
+      "loss": 0.6017,
+      "step": 5058
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6840612327982274,
+      "learning_rate": 7.922293579118165e-06,
+      "loss": 0.5505,
+      "step": 5059
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5578037891999774,
+      "learning_rate": 7.921456463687323e-06,
+      "loss": 0.5551,
+      "step": 5060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7035750119166656,
+      "learning_rate": 7.920619223899413e-06,
+      "loss": 0.5665,
+      "step": 5061
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7074648346030253,
+      "learning_rate": 7.919781859790076e-06,
+      "loss": 0.5221,
+      "step": 5062
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7985287647041702,
+      "learning_rate": 7.918944371394953e-06,
+      "loss": 0.5882,
+      "step": 5063
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6836268680802838,
+      "learning_rate": 7.9181067587497e-06,
+      "loss": 0.5696,
+      "step": 5064
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7989420764743823,
+      "learning_rate": 7.917269021889965e-06,
+      "loss": 0.5797,
+      "step": 5065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7147370856602557,
+      "learning_rate": 7.91643116085141e-06,
+      "loss": 0.565,
+      "step": 5066
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7860893362529873,
+      "learning_rate": 7.915593175669702e-06,
+      "loss": 0.601,
+      "step": 5067
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8811939473588392,
+      "learning_rate": 7.914755066380508e-06,
+      "loss": 0.6297,
+      "step": 5068
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6325432752600508,
+      "learning_rate": 7.913916833019503e-06,
+      "loss": 0.5523,
+      "step": 5069
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5665737980940468,
+      "learning_rate": 7.913078475622373e-06,
+      "loss": 0.4338,
+      "step": 5070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7085118896327416,
+      "learning_rate": 7.9122399942248e-06,
+      "loss": 0.608,
+      "step": 5071
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6413836376839173,
+      "learning_rate": 7.911401388862477e-06,
+      "loss": 0.5476,
+      "step": 5072
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7335134286632833,
+      "learning_rate": 7.9105626595711e-06,
+      "loss": 0.5112,
+      "step": 5073
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6778552221130795,
+      "learning_rate": 7.909723806386372e-06,
+      "loss": 0.5642,
+      "step": 5074
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6949316492523618,
+      "learning_rate": 7.908884829343998e-06,
+      "loss": 0.6151,
+      "step": 5075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6620510132905717,
+      "learning_rate": 7.908045728479694e-06,
+      "loss": 0.5053,
+      "step": 5076
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6588846465849989,
+      "learning_rate": 7.907206503829176e-06,
+      "loss": 0.5192,
+      "step": 5077
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8842365911384155,
+      "learning_rate": 7.906367155428168e-06,
+      "loss": 0.6151,
+      "step": 5078
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.628025438165786,
+      "learning_rate": 7.905527683312395e-06,
+      "loss": 0.524,
+      "step": 5079
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6083744230761006,
+      "learning_rate": 7.904688087517595e-06,
+      "loss": 0.5431,
+      "step": 5080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5776827949960666,
+      "learning_rate": 7.903848368079506e-06,
+      "loss": 0.505,
+      "step": 5081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5946324858724029,
+      "learning_rate": 7.903008525033868e-06,
+      "loss": 0.4874,
+      "step": 5082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7214488254666845,
+      "learning_rate": 7.902168558416436e-06,
+      "loss": 0.5664,
+      "step": 5083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7315511424239522,
+      "learning_rate": 7.901328468262963e-06,
+      "loss": 0.5936,
+      "step": 5084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6636904846832666,
+      "learning_rate": 7.900488254609209e-06,
+      "loss": 0.5117,
+      "step": 5085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6531652372518333,
+      "learning_rate": 7.899647917490938e-06,
+      "loss": 0.5689,
+      "step": 5086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9240937186441299,
+      "learning_rate": 7.89880745694392e-06,
+      "loss": 0.65,
+      "step": 5087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7957616863220984,
+      "learning_rate": 7.897966873003933e-06,
+      "loss": 0.6273,
+      "step": 5088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7467367534511393,
+      "learning_rate": 7.897126165706757e-06,
+      "loss": 0.5711,
+      "step": 5089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3916064669992472,
+      "learning_rate": 7.896285335088177e-06,
+      "loss": 0.6094,
+      "step": 5090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6232743246329217,
+      "learning_rate": 7.895444381183985e-06,
+      "loss": 0.5025,
+      "step": 5091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7756969175785325,
+      "learning_rate": 7.89460330402998e-06,
+      "loss": 0.6654,
+      "step": 5092
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7707337467045592,
+      "learning_rate": 7.89376210366196e-06,
+      "loss": 0.6015,
+      "step": 5093
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5844229986305767,
+      "learning_rate": 7.892920780115737e-06,
+      "loss": 0.5971,
+      "step": 5094
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7163947909411027,
+      "learning_rate": 7.892079333427118e-06,
+      "loss": 0.6132,
+      "step": 5095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5902806092067072,
+      "learning_rate": 7.891237763631925e-06,
+      "loss": 0.4839,
+      "step": 5096
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.68979041909195,
+      "learning_rate": 7.89039607076598e-06,
+      "loss": 0.6174,
+      "step": 5097
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8002207584875587,
+      "learning_rate": 7.889554254865111e-06,
+      "loss": 0.5945,
+      "step": 5098
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7854459286449994,
+      "learning_rate": 7.888712315965149e-06,
+      "loss": 0.6174,
+      "step": 5099
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7525764975305921,
+      "learning_rate": 7.887870254101935e-06,
+      "loss": 0.6494,
+      "step": 5100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6643851045720129,
+      "learning_rate": 7.887028069311313e-06,
+      "loss": 0.5931,
+      "step": 5101
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8465028869449369,
+      "learning_rate": 7.886185761629132e-06,
+      "loss": 0.6563,
+      "step": 5102
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8076228025908405,
+      "learning_rate": 7.885343331091247e-06,
+      "loss": 0.6046,
+      "step": 5103
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5874870140906882,
+      "learning_rate": 7.884500777733516e-06,
+      "loss": 0.5657,
+      "step": 5104
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.226635961562645,
+      "learning_rate": 7.883658101591804e-06,
+      "loss": 0.558,
+      "step": 5105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7058727457265803,
+      "learning_rate": 7.882815302701982e-06,
+      "loss": 0.5941,
+      "step": 5106
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5803580049785869,
+      "learning_rate": 7.881972381099925e-06,
+      "loss": 0.5172,
+      "step": 5107
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6040349374902021,
+      "learning_rate": 7.881129336821512e-06,
+      "loss": 0.5331,
+      "step": 5108
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6026742740070428,
+      "learning_rate": 7.88028616990263e-06,
+      "loss": 0.5429,
+      "step": 5109
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6927095988027014,
+      "learning_rate": 7.879442880379172e-06,
+      "loss": 0.6484,
+      "step": 5110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5754632919120132,
+      "learning_rate": 7.87859946828703e-06,
+      "loss": 0.5088,
+      "step": 5111
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5707306306612392,
+      "learning_rate": 7.877755933662108e-06,
+      "loss": 0.4958,
+      "step": 5112
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7392420938090287,
+      "learning_rate": 7.876912276540311e-06,
+      "loss": 0.6147,
+      "step": 5113
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5479266888669612,
+      "learning_rate": 7.876068496957552e-06,
+      "loss": 0.516,
+      "step": 5114
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.758498682855062,
+      "learning_rate": 7.875224594949748e-06,
+      "loss": 0.664,
+      "step": 5115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6984565023936016,
+      "learning_rate": 7.874380570552822e-06,
+      "loss": 0.5542,
+      "step": 5116
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5626514143291393,
+      "learning_rate": 7.8735364238027e-06,
+      "loss": 0.4988,
+      "step": 5117
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8381434701455139,
+      "learning_rate": 7.872692154735313e-06,
+      "loss": 0.6667,
+      "step": 5118
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8450080472542723,
+      "learning_rate": 7.871847763386602e-06,
+      "loss": 0.6469,
+      "step": 5119
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6069234026750302,
+      "learning_rate": 7.871003249792508e-06,
+      "loss": 0.5464,
+      "step": 5120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7521649567792072,
+      "learning_rate": 7.87015861398898e-06,
+      "loss": 0.5728,
+      "step": 5121
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8635092037279701,
+      "learning_rate": 7.869313856011974e-06,
+      "loss": 0.6384,
+      "step": 5122
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5564774881543907,
+      "learning_rate": 7.868468975897445e-06,
+      "loss": 0.5411,
+      "step": 5123
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5449304868682998,
+      "learning_rate": 7.867623973681358e-06,
+      "loss": 0.4851,
+      "step": 5124
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6184281164995215,
+      "learning_rate": 7.86677884939968e-06,
+      "loss": 0.531,
+      "step": 5125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8489258598200148,
+      "learning_rate": 7.865933603088389e-06,
+      "loss": 0.6572,
+      "step": 5126
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.638792438542576,
+      "learning_rate": 7.865088234783463e-06,
+      "loss": 0.5859,
+      "step": 5127
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7543736467193048,
+      "learning_rate": 7.864242744520886e-06,
+      "loss": 0.5805,
+      "step": 5128
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7625111433611972,
+      "learning_rate": 7.863397132336648e-06,
+      "loss": 0.6078,
+      "step": 5129
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7196372673758752,
+      "learning_rate": 7.862551398266744e-06,
+      "loss": 0.5841,
+      "step": 5130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8401073786095281,
+      "learning_rate": 7.861705542347175e-06,
+      "loss": 0.5828,
+      "step": 5131
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6065509265600705,
+      "learning_rate": 7.860859564613945e-06,
+      "loss": 0.5115,
+      "step": 5132
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0771813157619665,
+      "learning_rate": 7.860013465103065e-06,
+      "loss": 0.6201,
+      "step": 5133
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5939155001641759,
+      "learning_rate": 7.859167243850552e-06,
+      "loss": 0.4979,
+      "step": 5134
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7375112666833743,
+      "learning_rate": 7.858320900892427e-06,
+      "loss": 0.6139,
+      "step": 5135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6441646194030141,
+      "learning_rate": 7.857474436264714e-06,
+      "loss": 0.5202,
+      "step": 5136
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6054882863749236,
+      "learning_rate": 7.856627850003444e-06,
+      "loss": 0.5316,
+      "step": 5137
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8045266530316749,
+      "learning_rate": 7.855781142144658e-06,
+      "loss": 0.6269,
+      "step": 5138
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9468585915100367,
+      "learning_rate": 7.854934312724391e-06,
+      "loss": 0.6268,
+      "step": 5139
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6460564456698489,
+      "learning_rate": 7.854087361778698e-06,
+      "loss": 0.5551,
+      "step": 5140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7607520314008942,
+      "learning_rate": 7.853240289343621e-06,
+      "loss": 0.6268,
+      "step": 5141
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6359834973801624,
+      "learning_rate": 7.852393095455226e-06,
+      "loss": 0.5635,
+      "step": 5142
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9100654112663452,
+      "learning_rate": 7.851545780149571e-06,
+      "loss": 0.6246,
+      "step": 5143
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6760599372913728,
+      "learning_rate": 7.850698343462724e-06,
+      "loss": 0.5719,
+      "step": 5144
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7202926353213159,
+      "learning_rate": 7.849850785430759e-06,
+      "loss": 0.5632,
+      "step": 5145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7183561679175698,
+      "learning_rate": 7.849003106089754e-06,
+      "loss": 0.5408,
+      "step": 5146
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6055642801620112,
+      "learning_rate": 7.848155305475789e-06,
+      "loss": 0.5052,
+      "step": 5147
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7046992852981054,
+      "learning_rate": 7.847307383624953e-06,
+      "loss": 0.5302,
+      "step": 5148
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7820383122734619,
+      "learning_rate": 7.846459340573342e-06,
+      "loss": 0.6188,
+      "step": 5149
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7600596662095117,
+      "learning_rate": 7.845611176357054e-06,
+      "loss": 0.6185,
+      "step": 5150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6274596287492995,
+      "learning_rate": 7.844762891012188e-06,
+      "loss": 0.5299,
+      "step": 5151
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7672853534932366,
+      "learning_rate": 7.84391448457486e-06,
+      "loss": 0.629,
+      "step": 5152
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5741163666937225,
+      "learning_rate": 7.843065957081178e-06,
+      "loss": 0.5292,
+      "step": 5153
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8314494622019364,
+      "learning_rate": 7.842217308567264e-06,
+      "loss": 0.6659,
+      "step": 5154
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6011434329075702,
+      "learning_rate": 7.841368539069242e-06,
+      "loss": 0.5511,
+      "step": 5155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8361578191754074,
+      "learning_rate": 7.840519648623239e-06,
+      "loss": 0.5895,
+      "step": 5156
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5632250279890848,
+      "learning_rate": 7.839670637265394e-06,
+      "loss": 0.4852,
+      "step": 5157
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7246292297571776,
+      "learning_rate": 7.838821505031844e-06,
+      "loss": 0.6514,
+      "step": 5158
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7391546158511483,
+      "learning_rate": 7.837972251958732e-06,
+      "loss": 0.6161,
+      "step": 5159
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8163291036002391,
+      "learning_rate": 7.837122878082212e-06,
+      "loss": 0.6397,
+      "step": 5160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8075185919964452,
+      "learning_rate": 7.836273383438438e-06,
+      "loss": 0.6156,
+      "step": 5161
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7559379472919219,
+      "learning_rate": 7.835423768063567e-06,
+      "loss": 0.6047,
+      "step": 5162
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8136407199916558,
+      "learning_rate": 7.834574031993768e-06,
+      "loss": 0.6001,
+      "step": 5163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7023863094800844,
+      "learning_rate": 7.833724175265211e-06,
+      "loss": 0.578,
+      "step": 5164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6659585123110534,
+      "learning_rate": 7.83287419791407e-06,
+      "loss": 0.5229,
+      "step": 5165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6094398168476903,
+      "learning_rate": 7.832024099976528e-06,
+      "loss": 0.5229,
+      "step": 5166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6019145873911041,
+      "learning_rate": 7.83117388148877e-06,
+      "loss": 0.4999,
+      "step": 5167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8972452164358072,
+      "learning_rate": 7.830323542486986e-06,
+      "loss": 0.6029,
+      "step": 5168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.626960742856091,
+      "learning_rate": 7.829473083007375e-06,
+      "loss": 0.5823,
+      "step": 5169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7269685635796725,
+      "learning_rate": 7.828622503086136e-06,
+      "loss": 0.6062,
+      "step": 5170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7469406272220259,
+      "learning_rate": 7.827771802759474e-06,
+      "loss": 0.608,
+      "step": 5171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7433024029517992,
+      "learning_rate": 7.826920982063605e-06,
+      "loss": 0.5773,
+      "step": 5172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7168109881004537,
+      "learning_rate": 7.826070041034742e-06,
+      "loss": 0.5221,
+      "step": 5173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7240518256261765,
+      "learning_rate": 7.825218979709108e-06,
+      "loss": 0.5505,
+      "step": 5174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5734171472591354,
+      "learning_rate": 7.824367798122931e-06,
+      "loss": 0.5272,
+      "step": 5175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6215298058927083,
+      "learning_rate": 7.823516496312443e-06,
+      "loss": 0.5087,
+      "step": 5176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7002335222577787,
+      "learning_rate": 7.822665074313878e-06,
+      "loss": 0.5872,
+      "step": 5177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8213398502255651,
+      "learning_rate": 7.821813532163483e-06,
+      "loss": 0.6456,
+      "step": 5178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8025881365393628,
+      "learning_rate": 7.820961869897502e-06,
+      "loss": 0.6218,
+      "step": 5179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6121564975239244,
+      "learning_rate": 7.820110087552189e-06,
+      "loss": 0.5248,
+      "step": 5180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7328072190869244,
+      "learning_rate": 7.819258185163801e-06,
+      "loss": 0.5726,
+      "step": 5181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.585275707297132,
+      "learning_rate": 7.818406162768602e-06,
+      "loss": 0.5403,
+      "step": 5182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6375131315404875,
+      "learning_rate": 7.817554020402859e-06,
+      "loss": 0.5232,
+      "step": 5183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6944233175821881,
+      "learning_rate": 7.816701758102845e-06,
+      "loss": 0.5848,
+      "step": 5184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8608587649765883,
+      "learning_rate": 7.815849375904838e-06,
+      "loss": 0.6514,
+      "step": 5185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6201659343923066,
+      "learning_rate": 7.81499687384512e-06,
+      "loss": 0.5047,
+      "step": 5186
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7537271775227715,
+      "learning_rate": 7.814144251959981e-06,
+      "loss": 0.5925,
+      "step": 5187
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7287519655576999,
+      "learning_rate": 7.813291510285713e-06,
+      "loss": 0.5735,
+      "step": 5188
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6470465468464696,
+      "learning_rate": 7.812438648858617e-06,
+      "loss": 0.4743,
+      "step": 5189
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6886438333736695,
+      "learning_rate": 7.811585667714995e-06,
+      "loss": 0.5626,
+      "step": 5190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6946642431259499,
+      "learning_rate": 7.810732566891154e-06,
+      "loss": 0.6052,
+      "step": 5191
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6308731610498515,
+      "learning_rate": 7.809879346423413e-06,
+      "loss": 0.5125,
+      "step": 5192
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6106891296862297,
+      "learning_rate": 7.809026006348084e-06,
+      "loss": 0.5201,
+      "step": 5193
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5615808931381902,
+      "learning_rate": 7.808172546701494e-06,
+      "loss": 0.5032,
+      "step": 5194
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7863016380805363,
+      "learning_rate": 7.807318967519974e-06,
+      "loss": 0.6359,
+      "step": 5195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.745459832326678,
+      "learning_rate": 7.806465268839856e-06,
+      "loss": 0.5823,
+      "step": 5196
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8146838361068188,
+      "learning_rate": 7.80561145069748e-06,
+      "loss": 0.6344,
+      "step": 5197
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.727117847301932,
+      "learning_rate": 7.80475751312919e-06,
+      "loss": 0.5705,
+      "step": 5198
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5648858992023471,
+      "learning_rate": 7.803903456171335e-06,
+      "loss": 0.5066,
+      "step": 5199
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6611353478365266,
+      "learning_rate": 7.803049279860269e-06,
+      "loss": 0.5384,
+      "step": 5200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8299260810403722,
+      "learning_rate": 7.802194984232354e-06,
+      "loss": 0.6377,
+      "step": 5201
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5871918397856358,
+      "learning_rate": 7.801340569323951e-06,
+      "loss": 0.4882,
+      "step": 5202
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6599277193781682,
+      "learning_rate": 7.800486035171435e-06,
+      "loss": 0.5231,
+      "step": 5203
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8046535975450259,
+      "learning_rate": 7.799631381811176e-06,
+      "loss": 0.5779,
+      "step": 5204
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.701747700038936,
+      "learning_rate": 7.798776609279556e-06,
+      "loss": 0.5708,
+      "step": 5205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6359479503432651,
+      "learning_rate": 7.797921717612958e-06,
+      "loss": 0.5053,
+      "step": 5206
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6905023218957824,
+      "learning_rate": 7.797066706847774e-06,
+      "loss": 0.6112,
+      "step": 5207
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6848635570336745,
+      "learning_rate": 7.796211577020397e-06,
+      "loss": 0.5701,
+      "step": 5208
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7129518356470682,
+      "learning_rate": 7.79535632816723e-06,
+      "loss": 0.6381,
+      "step": 5209
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6438700220277365,
+      "learning_rate": 7.794500960324678e-06,
+      "loss": 0.5188,
+      "step": 5210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7216950014145777,
+      "learning_rate": 7.793645473529149e-06,
+      "loss": 0.5319,
+      "step": 5211
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8354405557326996,
+      "learning_rate": 7.792789867817058e-06,
+      "loss": 0.624,
+      "step": 5212
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7324552468722789,
+      "learning_rate": 7.791934143224829e-06,
+      "loss": 0.6057,
+      "step": 5213
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7308574461977455,
+      "learning_rate": 7.791078299788884e-06,
+      "loss": 0.5678,
+      "step": 5214
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5744071979513701,
+      "learning_rate": 7.790222337545654e-06,
+      "loss": 0.5048,
+      "step": 5215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6416332515529836,
+      "learning_rate": 7.789366256531576e-06,
+      "loss": 0.5444,
+      "step": 5216
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6933249241088075,
+      "learning_rate": 7.788510056783092e-06,
+      "loss": 0.6455,
+      "step": 5217
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8108136984915921,
+      "learning_rate": 7.787653738336643e-06,
+      "loss": 0.6444,
+      "step": 5218
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6048875555064881,
+      "learning_rate": 7.786797301228684e-06,
+      "loss": 0.5507,
+      "step": 5219
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.930254418294972,
+      "learning_rate": 7.785940745495668e-06,
+      "loss": 0.6532,
+      "step": 5220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7822429191860759,
+      "learning_rate": 7.785084071174058e-06,
+      "loss": 0.5916,
+      "step": 5221
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9170950010327478,
+      "learning_rate": 7.784227278300318e-06,
+      "loss": 0.6079,
+      "step": 5222
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.673604555362709,
+      "learning_rate": 7.783370366910922e-06,
+      "loss": 0.5987,
+      "step": 5223
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6986338181314327,
+      "learning_rate": 7.78251333704234e-06,
+      "loss": 0.6355,
+      "step": 5224
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6307382224520708,
+      "learning_rate": 7.781656188731062e-06,
+      "loss": 0.5691,
+      "step": 5225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6829051843925056,
+      "learning_rate": 7.780798922013565e-06,
+      "loss": 0.5189,
+      "step": 5226
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6749744785196069,
+      "learning_rate": 7.779941536926345e-06,
+      "loss": 0.5651,
+      "step": 5227
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6058614720841491,
+      "learning_rate": 7.779084033505899e-06,
+      "loss": 0.522,
+      "step": 5228
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8298931683925259,
+      "learning_rate": 7.778226411788725e-06,
+      "loss": 0.62,
+      "step": 5229
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7912802282173766,
+      "learning_rate": 7.77736867181133e-06,
+      "loss": 0.6445,
+      "step": 5230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8074130351241093,
+      "learning_rate": 7.776510813610225e-06,
+      "loss": 0.6185,
+      "step": 5231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7578963500252686,
+      "learning_rate": 7.77565283722193e-06,
+      "loss": 0.5821,
+      "step": 5232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6218647626778033,
+      "learning_rate": 7.774794742682961e-06,
+      "loss": 0.5173,
+      "step": 5233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6471351305066092,
+      "learning_rate": 7.773936530029849e-06,
+      "loss": 0.5417,
+      "step": 5234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7390504882646253,
+      "learning_rate": 7.77307819929912e-06,
+      "loss": 0.6398,
+      "step": 5235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7753625720932061,
+      "learning_rate": 7.772219750527316e-06,
+      "loss": 0.652,
+      "step": 5236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6519807511949876,
+      "learning_rate": 7.771361183750975e-06,
+      "loss": 0.5764,
+      "step": 5237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7810016907333998,
+      "learning_rate": 7.770502499006645e-06,
+      "loss": 0.6333,
+      "step": 5238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5838930966048808,
+      "learning_rate": 7.769643696330877e-06,
+      "loss": 0.5447,
+      "step": 5239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0001303086954352,
+      "learning_rate": 7.768784775760228e-06,
+      "loss": 0.6272,
+      "step": 5240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6931588109953511,
+      "learning_rate": 7.767925737331256e-06,
+      "loss": 0.5593,
+      "step": 5241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5662650751063849,
+      "learning_rate": 7.767066581080534e-06,
+      "loss": 0.4843,
+      "step": 5242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6550126948099457,
+      "learning_rate": 7.766207307044627e-06,
+      "loss": 0.5182,
+      "step": 5243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.750923788877791,
+      "learning_rate": 7.765347915260115e-06,
+      "loss": 0.6606,
+      "step": 5244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7441033530674466,
+      "learning_rate": 7.76448840576358e-06,
+      "loss": 0.635,
+      "step": 5245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7096215012666898,
+      "learning_rate": 7.763628778591607e-06,
+      "loss": 0.7049,
+      "step": 5246
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7676337863445605,
+      "learning_rate": 7.76276903378079e-06,
+      "loss": 0.6122,
+      "step": 5247
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8137413588366649,
+      "learning_rate": 7.761909171367722e-06,
+      "loss": 0.6652,
+      "step": 5248
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6437698150916662,
+      "learning_rate": 7.761049191389007e-06,
+      "loss": 0.5222,
+      "step": 5249
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5676579201559623,
+      "learning_rate": 7.76018909388125e-06,
+      "loss": 0.5295,
+      "step": 5250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5614919750815595,
+      "learning_rate": 7.759328878881065e-06,
+      "loss": 0.5247,
+      "step": 5251
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5382690537083782,
+      "learning_rate": 7.758468546425065e-06,
+      "loss": 0.5207,
+      "step": 5252
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6729442376850997,
+      "learning_rate": 7.757608096549876e-06,
+      "loss": 0.595,
+      "step": 5253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6365130074032272,
+      "learning_rate": 7.756747529292122e-06,
+      "loss": 0.5417,
+      "step": 5254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8748126834650067,
+      "learning_rate": 7.755886844688434e-06,
+      "loss": 0.6467,
+      "step": 5255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6791106755834856,
+      "learning_rate": 7.755026042775449e-06,
+      "loss": 0.4985,
+      "step": 5256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1160212712389737,
+      "learning_rate": 7.75416512358981e-06,
+      "loss": 0.6542,
+      "step": 5257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5903330924371895,
+      "learning_rate": 7.753304087168164e-06,
+      "loss": 0.5641,
+      "step": 5258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7014534647262111,
+      "learning_rate": 7.75244293354716e-06,
+      "loss": 0.5763,
+      "step": 5259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.578510819540876,
+      "learning_rate": 7.751581662763457e-06,
+      "loss": 0.5263,
+      "step": 5260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7657974706803282,
+      "learning_rate": 7.750720274853714e-06,
+      "loss": 0.5801,
+      "step": 5261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.705884157972561,
+      "learning_rate": 7.749858769854599e-06,
+      "loss": 0.5401,
+      "step": 5262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6155866238805996,
+      "learning_rate": 7.748997147802784e-06,
+      "loss": 0.5116,
+      "step": 5263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7409621896719885,
+      "learning_rate": 7.748135408734944e-06,
+      "loss": 0.5894,
+      "step": 5264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6367364129316485,
+      "learning_rate": 7.747273552687764e-06,
+      "loss": 0.5494,
+      "step": 5265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.846871606813435,
+      "learning_rate": 7.746411579697925e-06,
+      "loss": 0.6348,
+      "step": 5266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.788538756400709,
+      "learning_rate": 7.745549489802122e-06,
+      "loss": 0.5791,
+      "step": 5267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6446296980648142,
+      "learning_rate": 7.744687283037052e-06,
+      "loss": 0.5506,
+      "step": 5268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.79133393596402,
+      "learning_rate": 7.743824959439414e-06,
+      "loss": 0.6569,
+      "step": 5269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9501967627153795,
+      "learning_rate": 7.742962519045918e-06,
+      "loss": 0.6764,
+      "step": 5270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7381205333942157,
+      "learning_rate": 7.74209996189327e-06,
+      "loss": 0.5695,
+      "step": 5271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6540025867844753,
+      "learning_rate": 7.741237288018191e-06,
+      "loss": 0.5615,
+      "step": 5272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7181243194896826,
+      "learning_rate": 7.740374497457398e-06,
+      "loss": 0.6131,
+      "step": 5273
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6566134632869463,
+      "learning_rate": 7.739511590247621e-06,
+      "loss": 0.6021,
+      "step": 5274
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8969440704119592,
+      "learning_rate": 7.738648566425591e-06,
+      "loss": 0.5757,
+      "step": 5275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7802441121795158,
+      "learning_rate": 7.737785426028041e-06,
+      "loss": 0.6476,
+      "step": 5276
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.91875618050464,
+      "learning_rate": 7.736922169091716e-06,
+      "loss": 0.5961,
+      "step": 5277
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8098334712758746,
+      "learning_rate": 7.736058795653359e-06,
+      "loss": 0.6284,
+      "step": 5278
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7593980031617001,
+      "learning_rate": 7.73519530574972e-06,
+      "loss": 0.6243,
+      "step": 5279
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7599905609209325,
+      "learning_rate": 7.73433169941756e-06,
+      "loss": 0.5997,
+      "step": 5280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6565273382115473,
+      "learning_rate": 7.733467976693637e-06,
+      "loss": 0.5329,
+      "step": 5281
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5459231059576544,
+      "learning_rate": 7.732604137614718e-06,
+      "loss": 0.5042,
+      "step": 5282
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.731271369521869,
+      "learning_rate": 7.73174018221757e-06,
+      "loss": 0.6515,
+      "step": 5283
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6690327574379991,
+      "learning_rate": 7.730876110538975e-06,
+      "loss": 0.584,
+      "step": 5284
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1085067758957585,
+      "learning_rate": 7.73001192261571e-06,
+      "loss": 0.7003,
+      "step": 5285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5691674781479082,
+      "learning_rate": 7.72914761848456e-06,
+      "loss": 0.5211,
+      "step": 5286
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7834924281762433,
+      "learning_rate": 7.728283198182318e-06,
+      "loss": 0.6152,
+      "step": 5287
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8045769745747544,
+      "learning_rate": 7.727418661745778e-06,
+      "loss": 0.6364,
+      "step": 5288
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6718009128217338,
+      "learning_rate": 7.726554009211741e-06,
+      "loss": 0.591,
+      "step": 5289
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7679383615974319,
+      "learning_rate": 7.725689240617013e-06,
+      "loss": 0.6062,
+      "step": 5290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6527731247062422,
+      "learning_rate": 7.724824355998407e-06,
+      "loss": 0.5649,
+      "step": 5291
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6553288435708892,
+      "learning_rate": 7.723959355392731e-06,
+      "loss": 0.6194,
+      "step": 5292
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6506231333113452,
+      "learning_rate": 7.723094238836812e-06,
+      "loss": 0.5562,
+      "step": 5293
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.151922057213763,
+      "learning_rate": 7.722229006367472e-06,
+      "loss": 0.6436,
+      "step": 5294
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.628292377199441,
+      "learning_rate": 7.721363658021544e-06,
+      "loss": 0.5476,
+      "step": 5295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6171340045203727,
+      "learning_rate": 7.72049819383586e-06,
+      "loss": 0.5805,
+      "step": 5296
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7147657361764136,
+      "learning_rate": 7.719632613847261e-06,
+      "loss": 0.5796,
+      "step": 5297
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8015877735837139,
+      "learning_rate": 7.718766918092591e-06,
+      "loss": 0.5992,
+      "step": 5298
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7026833989226147,
+      "learning_rate": 7.717901106608702e-06,
+      "loss": 0.6074,
+      "step": 5299
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8114330970205671,
+      "learning_rate": 7.71703517943245e-06,
+      "loss": 0.4924,
+      "step": 5300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7083457757063296,
+      "learning_rate": 7.716169136600693e-06,
+      "loss": 0.6061,
+      "step": 5301
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5583808615003,
+      "learning_rate": 7.715302978150294e-06,
+      "loss": 0.5014,
+      "step": 5302
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5871350625935311,
+      "learning_rate": 7.714436704118125e-06,
+      "loss": 0.5239,
+      "step": 5303
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6299127194613465,
+      "learning_rate": 7.713570314541058e-06,
+      "loss": 0.553,
+      "step": 5304
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8586505503492021,
+      "learning_rate": 7.712703809455975e-06,
+      "loss": 0.5921,
+      "step": 5305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6091045261816805,
+      "learning_rate": 7.71183718889976e-06,
+      "loss": 0.5248,
+      "step": 5306
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6034704887543436,
+      "learning_rate": 7.7109704529093e-06,
+      "loss": 0.5276,
+      "step": 5307
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7722605894684905,
+      "learning_rate": 7.710103601521492e-06,
+      "loss": 0.6225,
+      "step": 5308
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6710929798497455,
+      "learning_rate": 7.709236634773236e-06,
+      "loss": 0.4731,
+      "step": 5309
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7655579462306464,
+      "learning_rate": 7.708369552701432e-06,
+      "loss": 0.5772,
+      "step": 5310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7879824673690861,
+      "learning_rate": 7.707502355342989e-06,
+      "loss": 0.627,
+      "step": 5311
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6507972696036635,
+      "learning_rate": 7.706635042734824e-06,
+      "loss": 0.5595,
+      "step": 5312
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6743730344430495,
+      "learning_rate": 7.705767614913854e-06,
+      "loss": 0.534,
+      "step": 5313
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6790101412040879,
+      "learning_rate": 7.704900071917005e-06,
+      "loss": 0.5905,
+      "step": 5314
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8545274723722545,
+      "learning_rate": 7.704032413781202e-06,
+      "loss": 0.6125,
+      "step": 5315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6313573737284246,
+      "learning_rate": 7.703164640543382e-06,
+      "loss": 0.5129,
+      "step": 5316
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6075125460045357,
+      "learning_rate": 7.702296752240477e-06,
+      "loss": 0.5553,
+      "step": 5317
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6638290762596978,
+      "learning_rate": 7.701428748909437e-06,
+      "loss": 0.5404,
+      "step": 5318
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6688444155868246,
+      "learning_rate": 7.700560630587207e-06,
+      "loss": 0.5494,
+      "step": 5319
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6266998309402686,
+      "learning_rate": 7.699692397310743e-06,
+      "loss": 0.5443,
+      "step": 5320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6244766913497802,
+      "learning_rate": 7.698824049117e-06,
+      "loss": 0.5316,
+      "step": 5321
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8014393174243583,
+      "learning_rate": 7.69795558604294e-06,
+      "loss": 0.6258,
+      "step": 5322
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7211660414928572,
+      "learning_rate": 7.697087008125532e-06,
+      "loss": 0.6665,
+      "step": 5323
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7058654858828981,
+      "learning_rate": 7.696218315401751e-06,
+      "loss": 0.5718,
+      "step": 5324
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6178785197188702,
+      "learning_rate": 7.69534950790857e-06,
+      "loss": 0.5029,
+      "step": 5325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6951047346040349,
+      "learning_rate": 7.694480585682975e-06,
+      "loss": 0.5487,
+      "step": 5326
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8982887178203129,
+      "learning_rate": 7.69361154876195e-06,
+      "loss": 0.6249,
+      "step": 5327
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6800590187037101,
+      "learning_rate": 7.69274239718249e-06,
+      "loss": 0.5264,
+      "step": 5328
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6003970713816201,
+      "learning_rate": 7.691873130981591e-06,
+      "loss": 0.5264,
+      "step": 5329
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6240141298785197,
+      "learning_rate": 7.691003750196256e-06,
+      "loss": 0.5584,
+      "step": 5330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9179250672137081,
+      "learning_rate": 7.690134254863488e-06,
+      "loss": 0.6387,
+      "step": 5331
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.58800512162918,
+      "learning_rate": 7.689264645020303e-06,
+      "loss": 0.5629,
+      "step": 5332
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5639766958796449,
+      "learning_rate": 7.688394920703715e-06,
+      "loss": 0.5339,
+      "step": 5333
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8565263335822307,
+      "learning_rate": 7.687525081950747e-06,
+      "loss": 0.599,
+      "step": 5334
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8002941713789701,
+      "learning_rate": 7.686655128798423e-06,
+      "loss": 0.5762,
+      "step": 5335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7013084372962467,
+      "learning_rate": 7.685785061283776e-06,
+      "loss": 0.5826,
+      "step": 5336
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.71366480717337,
+      "learning_rate": 7.684914879443841e-06,
+      "loss": 0.5865,
+      "step": 5337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5942434556786707,
+      "learning_rate": 7.684044583315661e-06,
+      "loss": 0.5539,
+      "step": 5338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8754726154923553,
+      "learning_rate": 7.68317417293628e-06,
+      "loss": 0.6342,
+      "step": 5339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7556984430068773,
+      "learning_rate": 7.682303648342746e-06,
+      "loss": 0.5984,
+      "step": 5340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6625352947072592,
+      "learning_rate": 7.681433009572119e-06,
+      "loss": 0.5482,
+      "step": 5341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.66197751487079,
+      "learning_rate": 7.680562256661457e-06,
+      "loss": 0.5302,
+      "step": 5342
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7109019654491981,
+      "learning_rate": 7.679691389647825e-06,
+      "loss": 0.5847,
+      "step": 5343
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7787371126541046,
+      "learning_rate": 7.678820408568295e-06,
+      "loss": 0.6185,
+      "step": 5344
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.613323840463627,
+      "learning_rate": 7.67794931345994e-06,
+      "loss": 0.5122,
+      "step": 5345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6187148388381738,
+      "learning_rate": 7.67707810435984e-06,
+      "loss": 0.5817,
+      "step": 5346
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6664703757482348,
+      "learning_rate": 7.67620678130508e-06,
+      "loss": 0.5617,
+      "step": 5347
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8225262527102246,
+      "learning_rate": 7.67533534433275e-06,
+      "loss": 0.6193,
+      "step": 5348
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6386208623915298,
+      "learning_rate": 7.674463793479943e-06,
+      "loss": 0.555,
+      "step": 5349
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5982119011369568,
+      "learning_rate": 7.673592128783759e-06,
+      "loss": 0.5243,
+      "step": 5350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6490819455678268,
+      "learning_rate": 7.6727203502813e-06,
+      "loss": 0.4916,
+      "step": 5351
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6128393412017639,
+      "learning_rate": 7.67184845800968e-06,
+      "loss": 0.5057,
+      "step": 5352
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6780731281205047,
+      "learning_rate": 7.670976452006004e-06,
+      "loss": 0.5384,
+      "step": 5353
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5690235257815663,
+      "learning_rate": 7.6701043323074e-06,
+      "loss": 0.4932,
+      "step": 5354
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7420962431359314,
+      "learning_rate": 7.669232098950985e-06,
+      "loss": 0.6022,
+      "step": 5355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.635443809239543,
+      "learning_rate": 7.66835975197389e-06,
+      "loss": 0.5794,
+      "step": 5356
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7707091924932612,
+      "learning_rate": 7.667487291413247e-06,
+      "loss": 0.616,
+      "step": 5357
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5705184601449362,
+      "learning_rate": 7.666614717306193e-06,
+      "loss": 0.5002,
+      "step": 5358
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1716942186392239,
+      "learning_rate": 7.665742029689874e-06,
+      "loss": 0.6027,
+      "step": 5359
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7184567838817191,
+      "learning_rate": 7.664869228601433e-06,
+      "loss": 0.5835,
+      "step": 5360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7024049582923059,
+      "learning_rate": 7.663996314078025e-06,
+      "loss": 0.5724,
+      "step": 5361
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.539286470647893,
+      "learning_rate": 7.663123286156809e-06,
+      "loss": 0.4751,
+      "step": 5362
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6091919738277533,
+      "learning_rate": 7.662250144874943e-06,
+      "loss": 0.5182,
+      "step": 5363
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.582008114127871,
+      "learning_rate": 7.661376890269595e-06,
+      "loss": 0.5199,
+      "step": 5364
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5617807601456263,
+      "learning_rate": 7.660503522377938e-06,
+      "loss": 0.5081,
+      "step": 5365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6430367415995056,
+      "learning_rate": 7.659630041237148e-06,
+      "loss": 0.5466,
+      "step": 5366
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6385091164378073,
+      "learning_rate": 7.658756446884405e-06,
+      "loss": 0.5704,
+      "step": 5367
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.682410731978004,
+      "learning_rate": 7.657882739356898e-06,
+      "loss": 0.5996,
+      "step": 5368
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7599238778128445,
+      "learning_rate": 7.657008918691816e-06,
+      "loss": 0.6487,
+      "step": 5369
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8529496451834723,
+      "learning_rate": 7.656134984926354e-06,
+      "loss": 0.6324,
+      "step": 5370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5668754506049037,
+      "learning_rate": 7.655260938097715e-06,
+      "loss": 0.5344,
+      "step": 5371
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6120986039999534,
+      "learning_rate": 7.654386778243102e-06,
+      "loss": 0.5122,
+      "step": 5372
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5789339225907278,
+      "learning_rate": 7.653512505399725e-06,
+      "loss": 0.5343,
+      "step": 5373
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6454722610257745,
+      "learning_rate": 7.6526381196048e-06,
+      "loss": 0.5546,
+      "step": 5374
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7044088838158508,
+      "learning_rate": 7.65176362089555e-06,
+      "loss": 0.6006,
+      "step": 5375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.581249987029883,
+      "learning_rate": 7.650889009309192e-06,
+      "loss": 0.512,
+      "step": 5376
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8576964304073412,
+      "learning_rate": 7.650014284882963e-06,
+      "loss": 0.6489,
+      "step": 5377
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.545865383601175,
+      "learning_rate": 7.649139447654093e-06,
+      "loss": 0.4896,
+      "step": 5378
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6026253881184362,
+      "learning_rate": 7.648264497659823e-06,
+      "loss": 0.5201,
+      "step": 5379
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.810654222136975,
+      "learning_rate": 7.647389434937396e-06,
+      "loss": 0.5675,
+      "step": 5380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6133869022233376,
+      "learning_rate": 7.64651425952406e-06,
+      "loss": 0.5084,
+      "step": 5381
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8385442460914531,
+      "learning_rate": 7.645638971457068e-06,
+      "loss": 0.5815,
+      "step": 5382
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5621602526736209,
+      "learning_rate": 7.64476357077368e-06,
+      "loss": 0.4776,
+      "step": 5383
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6911636496925291,
+      "learning_rate": 7.643888057511159e-06,
+      "loss": 0.5934,
+      "step": 5384
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6843306488745862,
+      "learning_rate": 7.643012431706771e-06,
+      "loss": 0.5266,
+      "step": 5385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7198955846421942,
+      "learning_rate": 7.64213669339779e-06,
+      "loss": 0.5934,
+      "step": 5386
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6491076186574289,
+      "learning_rate": 7.641260842621494e-06,
+      "loss": 0.5063,
+      "step": 5387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6641918581037313,
+      "learning_rate": 7.640384879415164e-06,
+      "loss": 0.5459,
+      "step": 5388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7564332359919856,
+      "learning_rate": 7.639508803816087e-06,
+      "loss": 0.6161,
+      "step": 5389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5403020534820326,
+      "learning_rate": 7.638632615861555e-06,
+      "loss": 0.4545,
+      "step": 5390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7666645707024677,
+      "learning_rate": 7.637756315588862e-06,
+      "loss": 0.6201,
+      "step": 5391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.835280670452697,
+      "learning_rate": 7.636879903035314e-06,
+      "loss": 0.617,
+      "step": 5392
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6937853179061539,
+      "learning_rate": 7.636003378238215e-06,
+      "loss": 0.5016,
+      "step": 5393
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6569690233668273,
+      "learning_rate": 7.635126741234873e-06,
+      "loss": 0.5547,
+      "step": 5394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7645363855007361,
+      "learning_rate": 7.63424999206261e-06,
+      "loss": 0.6731,
+      "step": 5395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5415710308418598,
+      "learning_rate": 7.633373130758741e-06,
+      "loss": 0.5246,
+      "step": 5396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7086320400014732,
+      "learning_rate": 7.632496157360591e-06,
+      "loss": 0.5988,
+      "step": 5397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5935657497173328,
+      "learning_rate": 7.631619071905494e-06,
+      "loss": 0.5018,
+      "step": 5398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5834521141203658,
+      "learning_rate": 7.630741874430782e-06,
+      "loss": 0.5276,
+      "step": 5399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5481030974166777,
+      "learning_rate": 7.629864564973798e-06,
+      "loss": 0.4916,
+      "step": 5400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8143837782426001,
+      "learning_rate": 7.62898714357188e-06,
+      "loss": 0.6332,
+      "step": 5401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2293795422243798,
+      "learning_rate": 7.62810961026238e-06,
+      "loss": 0.6667,
+      "step": 5402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7048513179537687,
+      "learning_rate": 7.627231965082653e-06,
+      "loss": 0.535,
+      "step": 5403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7083127999938686,
+      "learning_rate": 7.626354208070057e-06,
+      "loss": 0.5071,
+      "step": 5404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5967341527148564,
+      "learning_rate": 7.625476339261955e-06,
+      "loss": 0.5027,
+      "step": 5405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6622242437396073,
+      "learning_rate": 7.624598358695714e-06,
+      "loss": 0.5336,
+      "step": 5406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6955140117451355,
+      "learning_rate": 7.6237202664087095e-06,
+      "loss": 0.5751,
+      "step": 5407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7103838535428126,
+      "learning_rate": 7.622842062438317e-06,
+      "loss": 0.6078,
+      "step": 5408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7475216235717348,
+      "learning_rate": 7.621963746821919e-06,
+      "loss": 0.6063,
+      "step": 5409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6925293170683312,
+      "learning_rate": 7.621085319596904e-06,
+      "loss": 0.6021,
+      "step": 5410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.607102851080319,
+      "learning_rate": 7.62020678080066e-06,
+      "loss": 0.5005,
+      "step": 5411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6573127097147509,
+      "learning_rate": 7.619328130470589e-06,
+      "loss": 0.5143,
+      "step": 5412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.724357688680548,
+      "learning_rate": 7.618449368644087e-06,
+      "loss": 0.6107,
+      "step": 5413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6472703477079979,
+      "learning_rate": 7.617570495358565e-06,
+      "loss": 0.5169,
+      "step": 5414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7683889088151806,
+      "learning_rate": 7.61669151065143e-06,
+      "loss": 0.5863,
+      "step": 5415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5961739433201265,
+      "learning_rate": 7.615812414560102e-06,
+      "loss": 0.5223,
+      "step": 5416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7755467138983485,
+      "learning_rate": 7.614933207121996e-06,
+      "loss": 0.609,
+      "step": 5417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6213787758286397,
+      "learning_rate": 7.6140538883745394e-06,
+      "loss": 0.5373,
+      "step": 5418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7239157114677028,
+      "learning_rate": 7.613174458355164e-06,
+      "loss": 0.6376,
+      "step": 5419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8068751355867426,
+      "learning_rate": 7.612294917101302e-06,
+      "loss": 0.6776,
+      "step": 5420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7230239352917054,
+      "learning_rate": 7.611415264650392e-06,
+      "loss": 0.643,
+      "step": 5421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6644945599757978,
+      "learning_rate": 7.610535501039878e-06,
+      "loss": 0.6106,
+      "step": 5422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5683173879743686,
+      "learning_rate": 7.609655626307211e-06,
+      "loss": 0.5046,
+      "step": 5423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7109541936676621,
+      "learning_rate": 7.608775640489844e-06,
+      "loss": 0.6107,
+      "step": 5424
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6107410779022118,
+      "learning_rate": 7.607895543625233e-06,
+      "loss": 0.5724,
+      "step": 5425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7945096443773711,
+      "learning_rate": 7.6070153357508435e-06,
+      "loss": 0.6273,
+      "step": 5426
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7070430298179371,
+      "learning_rate": 7.6061350169041415e-06,
+      "loss": 0.5953,
+      "step": 5427
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7733949426303575,
+      "learning_rate": 7.605254587122599e-06,
+      "loss": 0.5823,
+      "step": 5428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7809194339482484,
+      "learning_rate": 7.604374046443697e-06,
+      "loss": 0.5006,
+      "step": 5429
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6251040077474513,
+      "learning_rate": 7.60349339490491e-06,
+      "loss": 0.5755,
+      "step": 5430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.795837385220522,
+      "learning_rate": 7.602612632543732e-06,
+      "loss": 0.5978,
+      "step": 5431
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.576667226610342,
+      "learning_rate": 7.60173175939765e-06,
+      "loss": 0.515,
+      "step": 5432
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6468163008548756,
+      "learning_rate": 7.60085077550416e-06,
+      "loss": 0.5676,
+      "step": 5433
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6780546892541153,
+      "learning_rate": 7.599969680900765e-06,
+      "loss": 0.5994,
+      "step": 5434
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7657552987026378,
+      "learning_rate": 7.5990884756249685e-06,
+      "loss": 0.613,
+      "step": 5435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0754882894038966,
+      "learning_rate": 7.5982071597142815e-06,
+      "loss": 0.6183,
+      "step": 5436
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5796398281401157,
+      "learning_rate": 7.597325733206218e-06,
+      "loss": 0.543,
+      "step": 5437
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7832954875863424,
+      "learning_rate": 7.596444196138297e-06,
+      "loss": 0.6057,
+      "step": 5438
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7701490375794444,
+      "learning_rate": 7.595562548548045e-06,
+      "loss": 0.6196,
+      "step": 5439
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7453484478446744,
+      "learning_rate": 7.5946807904729894e-06,
+      "loss": 0.5259,
+      "step": 5440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5822669148490245,
+      "learning_rate": 7.593798921950662e-06,
+      "loss": 0.5312,
+      "step": 5441
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7201675708314541,
+      "learning_rate": 7.592916943018604e-06,
+      "loss": 0.6383,
+      "step": 5442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7129321961162887,
+      "learning_rate": 7.592034853714359e-06,
+      "loss": 0.6,
+      "step": 5443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6956984087821456,
+      "learning_rate": 7.591152654075473e-06,
+      "loss": 0.6116,
+      "step": 5444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6734346629490349,
+      "learning_rate": 7.590270344139499e-06,
+      "loss": 0.5618,
+      "step": 5445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6328741817031723,
+      "learning_rate": 7.589387923943992e-06,
+      "loss": 0.5009,
+      "step": 5446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.301681239723456,
+      "learning_rate": 7.5885053935265155e-06,
+      "loss": 0.6442,
+      "step": 5447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8176790790323062,
+      "learning_rate": 7.587622752924637e-06,
+      "loss": 0.6091,
+      "step": 5448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7297976101130947,
+      "learning_rate": 7.586740002175928e-06,
+      "loss": 0.5673,
+      "step": 5449
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6506549382261865,
+      "learning_rate": 7.585857141317962e-06,
+      "loss": 0.5933,
+      "step": 5450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6608629796794794,
+      "learning_rate": 7.584974170388319e-06,
+      "loss": 0.5372,
+      "step": 5451
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6446043855865727,
+      "learning_rate": 7.584091089424589e-06,
+      "loss": 0.5698,
+      "step": 5452
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6616059324009207,
+      "learning_rate": 7.583207898464356e-06,
+      "loss": 0.5118,
+      "step": 5453
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6768717887359736,
+      "learning_rate": 7.582324597545219e-06,
+      "loss": 0.5281,
+      "step": 5454
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7068338364357585,
+      "learning_rate": 7.581441186704776e-06,
+      "loss": 0.5516,
+      "step": 5455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5690980398220856,
+      "learning_rate": 7.580557665980631e-06,
+      "loss": 0.5194,
+      "step": 5456
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.625019399391991,
+      "learning_rate": 7.579674035410391e-06,
+      "loss": 0.4861,
+      "step": 5457
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6799176783698669,
+      "learning_rate": 7.5787902950316725e-06,
+      "loss": 0.5666,
+      "step": 5458
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.708875205693146,
+      "learning_rate": 7.577906444882091e-06,
+      "loss": 0.6049,
+      "step": 5459
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5581560808860451,
+      "learning_rate": 7.5770224849992705e-06,
+      "loss": 0.5692,
+      "step": 5460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.349474955178804,
+      "learning_rate": 7.576138415420838e-06,
+      "loss": 0.6313,
+      "step": 5461
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6496242085299923,
+      "learning_rate": 7.575254236184424e-06,
+      "loss": 0.5417,
+      "step": 5462
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7364391519847628,
+      "learning_rate": 7.5743699473276664e-06,
+      "loss": 0.6295,
+      "step": 5463
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9249468560003917,
+      "learning_rate": 7.573485548888209e-06,
+      "loss": 0.6309,
+      "step": 5464
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7308284489970758,
+      "learning_rate": 7.572601040903693e-06,
+      "loss": 0.5864,
+      "step": 5465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6646316896812953,
+      "learning_rate": 7.571716423411772e-06,
+      "loss": 0.5318,
+      "step": 5466
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7871039854942252,
+      "learning_rate": 7.570831696450101e-06,
+      "loss": 0.6606,
+      "step": 5467
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5777473358378997,
+      "learning_rate": 7.569946860056341e-06,
+      "loss": 0.5046,
+      "step": 5468
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6865903037439381,
+      "learning_rate": 7.569061914268154e-06,
+      "loss": 0.5421,
+      "step": 5469
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6264948735094058,
+      "learning_rate": 7.568176859123212e-06,
+      "loss": 0.5668,
+      "step": 5470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8757852540791926,
+      "learning_rate": 7.5672916946591866e-06,
+      "loss": 0.6485,
+      "step": 5471
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7127381619652646,
+      "learning_rate": 7.566406420913759e-06,
+      "loss": 0.6035,
+      "step": 5472
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7086887400407788,
+      "learning_rate": 7.565521037924612e-06,
+      "loss": 0.5203,
+      "step": 5473
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7432649571650681,
+      "learning_rate": 7.564635545729431e-06,
+      "loss": 0.59,
+      "step": 5474
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.632263530343266,
+      "learning_rate": 7.563749944365909e-06,
+      "loss": 0.5628,
+      "step": 5475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7524947352704826,
+      "learning_rate": 7.562864233871746e-06,
+      "loss": 0.534,
+      "step": 5476
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5442806979374378,
+      "learning_rate": 7.561978414284643e-06,
+      "loss": 0.5033,
+      "step": 5477
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.731510330305729,
+      "learning_rate": 7.561092485642305e-06,
+      "loss": 0.6024,
+      "step": 5478
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6656624046368786,
+      "learning_rate": 7.560206447982443e-06,
+      "loss": 0.5929,
+      "step": 5479
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7745102302614559,
+      "learning_rate": 7.559320301342775e-06,
+      "loss": 0.5565,
+      "step": 5480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6797113051754411,
+      "learning_rate": 7.55843404576102e-06,
+      "loss": 0.5854,
+      "step": 5481
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7466460173279136,
+      "learning_rate": 7.557547681274903e-06,
+      "loss": 0.6156,
+      "step": 5482
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7288466337400398,
+      "learning_rate": 7.556661207922156e-06,
+      "loss": 0.5199,
+      "step": 5483
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0119325422624517,
+      "learning_rate": 7.555774625740509e-06,
+      "loss": 0.649,
+      "step": 5484
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7638263102320347,
+      "learning_rate": 7.554887934767703e-06,
+      "loss": 0.6565,
+      "step": 5485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7065218600244592,
+      "learning_rate": 7.554001135041482e-06,
+      "loss": 0.5312,
+      "step": 5486
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6434874856143814,
+      "learning_rate": 7.553114226599595e-06,
+      "loss": 0.5571,
+      "step": 5487
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6323723664848917,
+      "learning_rate": 7.552227209479794e-06,
+      "loss": 0.5305,
+      "step": 5488
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7267461648954927,
+      "learning_rate": 7.5513400837198355e-06,
+      "loss": 0.5722,
+      "step": 5489
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6671882548083844,
+      "learning_rate": 7.550452849357484e-06,
+      "loss": 0.6023,
+      "step": 5490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6674776173780151,
+      "learning_rate": 7.5495655064305025e-06,
+      "loss": 0.5253,
+      "step": 5491
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.746430621867441,
+      "learning_rate": 7.548678054976666e-06,
+      "loss": 0.6356,
+      "step": 5492
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6746456754429112,
+      "learning_rate": 7.54779049503375e-06,
+      "loss": 0.5319,
+      "step": 5493
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5909693788870325,
+      "learning_rate": 7.546902826639533e-06,
+      "loss": 0.5263,
+      "step": 5494
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6822441732105623,
+      "learning_rate": 7.546015049831802e-06,
+      "loss": 0.5531,
+      "step": 5495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8611769642948816,
+      "learning_rate": 7.545127164648347e-06,
+      "loss": 0.6259,
+      "step": 5496
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6240254873315566,
+      "learning_rate": 7.544239171126959e-06,
+      "loss": 0.5498,
+      "step": 5497
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6247896324463584,
+      "learning_rate": 7.5433510693054425e-06,
+      "loss": 0.56,
+      "step": 5498
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8005416297048783,
+      "learning_rate": 7.542462859221597e-06,
+      "loss": 0.6389,
+      "step": 5499
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5355148967485481,
+      "learning_rate": 7.541574540913232e-06,
+      "loss": 0.52,
+      "step": 5500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7439451661989224,
+      "learning_rate": 7.5406861144181606e-06,
+      "loss": 0.5985,
+      "step": 5501
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9114322341961698,
+      "learning_rate": 7.539797579774201e-06,
+      "loss": 0.6614,
+      "step": 5502
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7011849852364572,
+      "learning_rate": 7.538908937019174e-06,
+      "loss": 0.6,
+      "step": 5503
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5877618936349194,
+      "learning_rate": 7.538020186190907e-06,
+      "loss": 0.5305,
+      "step": 5504
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7128670619590266,
+      "learning_rate": 7.537131327327229e-06,
+      "loss": 0.5942,
+      "step": 5505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6583241096129865,
+      "learning_rate": 7.536242360465979e-06,
+      "loss": 0.5493,
+      "step": 5506
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7975469748019888,
+      "learning_rate": 7.535353285644998e-06,
+      "loss": 0.627,
+      "step": 5507
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7146920127920903,
+      "learning_rate": 7.534464102902129e-06,
+      "loss": 0.6266,
+      "step": 5508
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7334534430237059,
+      "learning_rate": 7.53357481227522e-06,
+      "loss": 0.5772,
+      "step": 5509
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9073657062311751,
+      "learning_rate": 7.53268541380213e-06,
+      "loss": 0.6741,
+      "step": 5510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6777630299392265,
+      "learning_rate": 7.531795907520714e-06,
+      "loss": 0.5558,
+      "step": 5511
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6457095570061134,
+      "learning_rate": 7.530906293468837e-06,
+      "loss": 0.5385,
+      "step": 5512
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6152226963370656,
+      "learning_rate": 7.530016571684366e-06,
+      "loss": 0.5043,
+      "step": 5513
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.642558496011311,
+      "learning_rate": 7.529126742205176e-06,
+      "loss": 0.5459,
+      "step": 5514
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5746935789627408,
+      "learning_rate": 7.5282368050691405e-06,
+      "loss": 0.555,
+      "step": 5515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6245333402941172,
+      "learning_rate": 7.527346760314145e-06,
+      "loss": 0.5692,
+      "step": 5516
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6533041992053598,
+      "learning_rate": 7.526456607978072e-06,
+      "loss": 0.5658,
+      "step": 5517
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7423793547101148,
+      "learning_rate": 7.5255663480988165e-06,
+      "loss": 0.5703,
+      "step": 5518
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5864841659833234,
+      "learning_rate": 7.524675980714272e-06,
+      "loss": 0.5449,
+      "step": 5519
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6568467936774076,
+      "learning_rate": 7.523785505862339e-06,
+      "loss": 0.4968,
+      "step": 5520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5663009235343691,
+      "learning_rate": 7.5228949235809205e-06,
+      "loss": 0.4981,
+      "step": 5521
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5477631262774487,
+      "learning_rate": 7.522004233907927e-06,
+      "loss": 0.5194,
+      "step": 5522
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.767181909781016,
+      "learning_rate": 7.521113436881273e-06,
+      "loss": 0.5831,
+      "step": 5523
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.93607736525853,
+      "learning_rate": 7.520222532538877e-06,
+      "loss": 0.5758,
+      "step": 5524
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7008607764510454,
+      "learning_rate": 7.519331520918662e-06,
+      "loss": 0.5337,
+      "step": 5525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7891733069469153,
+      "learning_rate": 7.518440402058553e-06,
+      "loss": 0.641,
+      "step": 5526
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8685535532634578,
+      "learning_rate": 7.517549175996485e-06,
+      "loss": 0.6258,
+      "step": 5527
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.755134567623352,
+      "learning_rate": 7.5166578427703914e-06,
+      "loss": 0.637,
+      "step": 5528
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7420557935642936,
+      "learning_rate": 7.5157664024182174e-06,
+      "loss": 0.5485,
+      "step": 5529
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.59662029397398,
+      "learning_rate": 7.5148748549779075e-06,
+      "loss": 0.5874,
+      "step": 5530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.693248989669816,
+      "learning_rate": 7.51398320048741e-06,
+      "loss": 0.556,
+      "step": 5531
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6317669816606988,
+      "learning_rate": 7.513091438984682e-06,
+      "loss": 0.5348,
+      "step": 5532
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7504080633376384,
+      "learning_rate": 7.5121995705076825e-06,
+      "loss": 0.5211,
+      "step": 5533
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5733965546187132,
+      "learning_rate": 7.511307595094375e-06,
+      "loss": 0.4929,
+      "step": 5534
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7869138745795831,
+      "learning_rate": 7.5104155127827275e-06,
+      "loss": 0.5317,
+      "step": 5535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7438322500729675,
+      "learning_rate": 7.509523323610713e-06,
+      "loss": 0.6496,
+      "step": 5536
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.577448630036473,
+      "learning_rate": 7.508631027616311e-06,
+      "loss": 0.5282,
+      "step": 5537
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6927512837277563,
+      "learning_rate": 7.507738624837502e-06,
+      "loss": 0.6542,
+      "step": 5538
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7953838617113616,
+      "learning_rate": 7.506846115312274e-06,
+      "loss": 0.634,
+      "step": 5539
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6284276738644262,
+      "learning_rate": 7.505953499078618e-06,
+      "loss": 0.5649,
+      "step": 5540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7195176283967624,
+      "learning_rate": 7.505060776174529e-06,
+      "loss": 0.6298,
+      "step": 5541
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7861746368351229,
+      "learning_rate": 7.504167946638007e-06,
+      "loss": 0.5821,
+      "step": 5542
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7098625392169347,
+      "learning_rate": 7.503275010507058e-06,
+      "loss": 0.5195,
+      "step": 5543
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7370578608999878,
+      "learning_rate": 7.502381967819692e-06,
+      "loss": 0.6558,
+      "step": 5544
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6854933639347381,
+      "learning_rate": 7.501488818613921e-06,
+      "loss": 0.5144,
+      "step": 5545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6374023047684221,
+      "learning_rate": 7.500595562927764e-06,
+      "loss": 0.5638,
+      "step": 5546
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7846165899240831,
+      "learning_rate": 7.499702200799246e-06,
+      "loss": 0.602,
+      "step": 5547
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7775624447052985,
+      "learning_rate": 7.4988087322663936e-06,
+      "loss": 0.5535,
+      "step": 5548
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.86649198116926,
+      "learning_rate": 7.497915157367237e-06,
+      "loss": 0.6449,
+      "step": 5549
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6199681158813892,
+      "learning_rate": 7.497021476139816e-06,
+      "loss": 0.5017,
+      "step": 5550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5509487381920122,
+      "learning_rate": 7.496127688622169e-06,
+      "loss": 0.5233,
+      "step": 5551
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5544781377612417,
+      "learning_rate": 7.495233794852342e-06,
+      "loss": 0.5086,
+      "step": 5552
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5923297570546956,
+      "learning_rate": 7.494339794868388e-06,
+      "loss": 0.5406,
+      "step": 5553
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7484659934821983,
+      "learning_rate": 7.493445688708358e-06,
+      "loss": 0.5456,
+      "step": 5554
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6575814254828445,
+      "learning_rate": 7.492551476410314e-06,
+      "loss": 0.5076,
+      "step": 5555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6520039793260324,
+      "learning_rate": 7.491657158012319e-06,
+      "loss": 0.4847,
+      "step": 5556
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8301138512520254,
+      "learning_rate": 7.4907627335524405e-06,
+      "loss": 0.6799,
+      "step": 5557
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7741470180064739,
+      "learning_rate": 7.489868203068752e-06,
+      "loss": 0.5918,
+      "step": 5558
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7109115392808868,
+      "learning_rate": 7.488973566599329e-06,
+      "loss": 0.6069,
+      "step": 5559
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.767178787813936,
+      "learning_rate": 7.488078824182257e-06,
+      "loss": 0.576,
+      "step": 5560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.581246501876493,
+      "learning_rate": 7.4871839758556185e-06,
+      "loss": 0.5452,
+      "step": 5561
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8460346470153032,
+      "learning_rate": 7.4862890216575065e-06,
+      "loss": 0.6142,
+      "step": 5562
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8801084456986683,
+      "learning_rate": 7.4853939616260174e-06,
+      "loss": 0.5992,
+      "step": 5563
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7579932606957567,
+      "learning_rate": 7.4844987957992485e-06,
+      "loss": 0.6464,
+      "step": 5564
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6864144690084117,
+      "learning_rate": 7.483603524215303e-06,
+      "loss": 0.5825,
+      "step": 5565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8509339686622452,
+      "learning_rate": 7.482708146912295e-06,
+      "loss": 0.6108,
+      "step": 5566
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.555737422673903,
+      "learning_rate": 7.481812663928334e-06,
+      "loss": 0.5675,
+      "step": 5567
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7807117691246142,
+      "learning_rate": 7.4809170753015395e-06,
+      "loss": 0.6354,
+      "step": 5568
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6876108956454414,
+      "learning_rate": 7.480021381070032e-06,
+      "loss": 0.5911,
+      "step": 5569
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5522113252313763,
+      "learning_rate": 7.479125581271939e-06,
+      "loss": 0.4529,
+      "step": 5570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5929854641676827,
+      "learning_rate": 7.478229675945392e-06,
+      "loss": 0.4925,
+      "step": 5571
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7239031681381652,
+      "learning_rate": 7.477333665128526e-06,
+      "loss": 0.6061,
+      "step": 5572
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6463245893762213,
+      "learning_rate": 7.4764375488594855e-06,
+      "loss": 0.5464,
+      "step": 5573
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7083116479145974,
+      "learning_rate": 7.47554132717641e-06,
+      "loss": 0.5667,
+      "step": 5574
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8719949481149875,
+      "learning_rate": 7.474645000117451e-06,
+      "loss": 0.6517,
+      "step": 5575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5770683532298139,
+      "learning_rate": 7.473748567720762e-06,
+      "loss": 0.5132,
+      "step": 5576
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7199319889513051,
+      "learning_rate": 7.472852030024501e-06,
+      "loss": 0.5643,
+      "step": 5577
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6908856397956292,
+      "learning_rate": 7.4719553870668324e-06,
+      "loss": 0.5836,
+      "step": 5578
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6841647098460729,
+      "learning_rate": 7.4710586388859215e-06,
+      "loss": 0.6108,
+      "step": 5579
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5849911087359764,
+      "learning_rate": 7.470161785519942e-06,
+      "loss": 0.5901,
+      "step": 5580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6602073844857029,
+      "learning_rate": 7.469264827007068e-06,
+      "loss": 0.5685,
+      "step": 5581
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6523554592045353,
+      "learning_rate": 7.46836776338548e-06,
+      "loss": 0.5861,
+      "step": 5582
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6890021902397042,
+      "learning_rate": 7.467470594693364e-06,
+      "loss": 0.5016,
+      "step": 5583
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6401967613016705,
+      "learning_rate": 7.466573320968912e-06,
+      "loss": 0.508,
+      "step": 5584
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6570991772661504,
+      "learning_rate": 7.465675942250314e-06,
+      "loss": 0.5759,
+      "step": 5585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0630567084386817,
+      "learning_rate": 7.464778458575771e-06,
+      "loss": 0.6719,
+      "step": 5586
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6374844016034613,
+      "learning_rate": 7.4638808699834855e-06,
+      "loss": 0.5183,
+      "step": 5587
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5731688138145928,
+      "learning_rate": 7.462983176511663e-06,
+      "loss": 0.4934,
+      "step": 5588
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9017871195801972,
+      "learning_rate": 7.462085378198519e-06,
+      "loss": 0.5813,
+      "step": 5589
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6838869058084027,
+      "learning_rate": 7.4611874750822675e-06,
+      "loss": 0.5739,
+      "step": 5590
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7481473094848149,
+      "learning_rate": 7.460289467201129e-06,
+      "loss": 0.6227,
+      "step": 5591
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7552485024500268,
+      "learning_rate": 7.459391354593333e-06,
+      "loss": 0.5599,
+      "step": 5592
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6282533925896012,
+      "learning_rate": 7.458493137297103e-06,
+      "loss": 0.5235,
+      "step": 5593
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6166800626199566,
+      "learning_rate": 7.457594815350678e-06,
+      "loss": 0.5384,
+      "step": 5594
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7920820713629735,
+      "learning_rate": 7.456696388792295e-06,
+      "loss": 0.5468,
+      "step": 5595
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7346765132397483,
+      "learning_rate": 7.455797857660196e-06,
+      "loss": 0.536,
+      "step": 5596
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6430437289013298,
+      "learning_rate": 7.454899221992632e-06,
+      "loss": 0.4796,
+      "step": 5597
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6721675578508051,
+      "learning_rate": 7.454000481827851e-06,
+      "loss": 0.5873,
+      "step": 5598
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6096008981029728,
+      "learning_rate": 7.453101637204111e-06,
+      "loss": 0.5323,
+      "step": 5599
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5925943527999598,
+      "learning_rate": 7.452202688159674e-06,
+      "loss": 0.5041,
+      "step": 5600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5728460009147941,
+      "learning_rate": 7.451303634732805e-06,
+      "loss": 0.5218,
+      "step": 5601
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6973336160605523,
+      "learning_rate": 7.450404476961773e-06,
+      "loss": 0.6362,
+      "step": 5602
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7958766218236931,
+      "learning_rate": 7.449505214884853e-06,
+      "loss": 0.6399,
+      "step": 5603
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6441371673883424,
+      "learning_rate": 7.448605848540324e-06,
+      "loss": 0.5445,
+      "step": 5604
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5541922008080312,
+      "learning_rate": 7.447706377966469e-06,
+      "loss": 0.4974,
+      "step": 5605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7344791006261536,
+      "learning_rate": 7.446806803201574e-06,
+      "loss": 0.6002,
+      "step": 5606
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7050838857584212,
+      "learning_rate": 7.445907124283933e-06,
+      "loss": 0.5986,
+      "step": 5607
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7662848114017572,
+      "learning_rate": 7.445007341251841e-06,
+      "loss": 0.6198,
+      "step": 5608
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9430902788951556,
+      "learning_rate": 7.444107454143601e-06,
+      "loss": 0.6372,
+      "step": 5609
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7552672058105092,
+      "learning_rate": 7.443207462997515e-06,
+      "loss": 0.6099,
+      "step": 5610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6819261497740676,
+      "learning_rate": 7.442307367851897e-06,
+      "loss": 0.55,
+      "step": 5611
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7249374279700982,
+      "learning_rate": 7.441407168745056e-06,
+      "loss": 0.5955,
+      "step": 5612
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6992416001775748,
+      "learning_rate": 7.440506865715316e-06,
+      "loss": 0.5949,
+      "step": 5613
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6546698374596573,
+      "learning_rate": 7.439606458800995e-06,
+      "loss": 0.4977,
+      "step": 5614
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7400243103854021,
+      "learning_rate": 7.438705948040426e-06,
+      "loss": 0.641,
+      "step": 5615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5775259899719023,
+      "learning_rate": 7.4378053334719345e-06,
+      "loss": 0.5262,
+      "step": 5616
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5750151254313061,
+      "learning_rate": 7.436904615133862e-06,
+      "loss": 0.4882,
+      "step": 5617
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7503319692774794,
+      "learning_rate": 7.436003793064548e-06,
+      "loss": 0.5577,
+      "step": 5618
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.577851017085559,
+      "learning_rate": 7.435102867302335e-06,
+      "loss": 0.5084,
+      "step": 5619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7988919347178581,
+      "learning_rate": 7.434201837885576e-06,
+      "loss": 0.6207,
+      "step": 5620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6414598371822323,
+      "learning_rate": 7.433300704852622e-06,
+      "loss": 0.5151,
+      "step": 5621
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6155064305103313,
+      "learning_rate": 7.432399468241833e-06,
+      "loss": 0.5176,
+      "step": 5622
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6295500286375426,
+      "learning_rate": 7.431498128091572e-06,
+      "loss": 0.5537,
+      "step": 5623
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8794832337445059,
+      "learning_rate": 7.4305966844402055e-06,
+      "loss": 0.6861,
+      "step": 5624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7617276757271866,
+      "learning_rate": 7.429695137326105e-06,
+      "loss": 0.642,
+      "step": 5625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.597706565973292,
+      "learning_rate": 7.428793486787648e-06,
+      "loss": 0.5264,
+      "step": 5626
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6784040452151213,
+      "learning_rate": 7.427891732863212e-06,
+      "loss": 0.5605,
+      "step": 5627
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6174673738110383,
+      "learning_rate": 7.426989875591183e-06,
+      "loss": 0.5353,
+      "step": 5628
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6309831888077089,
+      "learning_rate": 7.426087915009952e-06,
+      "loss": 0.5295,
+      "step": 5629
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5773689344234675,
+      "learning_rate": 7.42518585115791e-06,
+      "loss": 0.5117,
+      "step": 5630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6580223975721441,
+      "learning_rate": 7.424283684073456e-06,
+      "loss": 0.5813,
+      "step": 5631
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7694103118508562,
+      "learning_rate": 7.423381413794995e-06,
+      "loss": 0.6048,
+      "step": 5632
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7798427890138921,
+      "learning_rate": 7.422479040360928e-06,
+      "loss": 0.593,
+      "step": 5633
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5912828207632826,
+      "learning_rate": 7.42157656380967e-06,
+      "loss": 0.476,
+      "step": 5634
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6336847705950144,
+      "learning_rate": 7.4206739841796375e-06,
+      "loss": 0.4985,
+      "step": 5635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8028719077312637,
+      "learning_rate": 7.419771301509248e-06,
+      "loss": 0.632,
+      "step": 5636
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.667322966161425,
+      "learning_rate": 7.418868515836927e-06,
+      "loss": 0.5008,
+      "step": 5637
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5994255309463825,
+      "learning_rate": 7.4179656272011025e-06,
+      "loss": 0.5302,
+      "step": 5638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7060310675464113,
+      "learning_rate": 7.417062635640207e-06,
+      "loss": 0.5395,
+      "step": 5639
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6671501879542686,
+      "learning_rate": 7.4161595411926804e-06,
+      "loss": 0.5724,
+      "step": 5640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6086150095679586,
+      "learning_rate": 7.415256343896963e-06,
+      "loss": 0.5083,
+      "step": 5641
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.67659508707042,
+      "learning_rate": 7.4143530437915e-06,
+      "loss": 0.53,
+      "step": 5642
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5930453102751836,
+      "learning_rate": 7.413449640914744e-06,
+      "loss": 0.5051,
+      "step": 5643
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6773802712835152,
+      "learning_rate": 7.4125461353051495e-06,
+      "loss": 0.6126,
+      "step": 5644
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.638229018001181,
+      "learning_rate": 7.411642527001174e-06,
+      "loss": 0.53,
+      "step": 5645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5765617116730168,
+      "learning_rate": 7.410738816041283e-06,
+      "loss": 0.5203,
+      "step": 5646
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7021036191079978,
+      "learning_rate": 7.409835002463946e-06,
+      "loss": 0.56,
+      "step": 5647
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6326871301716849,
+      "learning_rate": 7.408931086307631e-06,
+      "loss": 0.5378,
+      "step": 5648
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6308750172678016,
+      "learning_rate": 7.408027067610819e-06,
+      "loss": 0.5477,
+      "step": 5649
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6086451695692009,
+      "learning_rate": 7.40712294641199e-06,
+      "loss": 0.5381,
+      "step": 5650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6221308692310638,
+      "learning_rate": 7.406218722749629e-06,
+      "loss": 0.548,
+      "step": 5651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7823893248720007,
+      "learning_rate": 7.405314396662224e-06,
+      "loss": 0.5919,
+      "step": 5652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5910429812763102,
+      "learning_rate": 7.404409968188274e-06,
+      "loss": 0.4926,
+      "step": 5653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.845404297963058,
+      "learning_rate": 7.403505437366274e-06,
+      "loss": 0.6041,
+      "step": 5654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6082515530765161,
+      "learning_rate": 7.40260080423473e-06,
+      "loss": 0.568,
+      "step": 5655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5486652756234054,
+      "learning_rate": 7.401696068832145e-06,
+      "loss": 0.4863,
+      "step": 5656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6023368850152669,
+      "learning_rate": 7.400791231197034e-06,
+      "loss": 0.5219,
+      "step": 5657
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6377601372870552,
+      "learning_rate": 7.399886291367913e-06,
+      "loss": 0.5574,
+      "step": 5658
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7918680740193286,
+      "learning_rate": 7.398981249383299e-06,
+      "loss": 0.5839,
+      "step": 5659
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5790747823396108,
+      "learning_rate": 7.398076105281722e-06,
+      "loss": 0.5479,
+      "step": 5660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7643797573132838,
+      "learning_rate": 7.397170859101705e-06,
+      "loss": 0.6538,
+      "step": 5661
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6298632808110909,
+      "learning_rate": 7.396265510881788e-06,
+      "loss": 0.5053,
+      "step": 5662
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6966163507054176,
+      "learning_rate": 7.3953600606605035e-06,
+      "loss": 0.5453,
+      "step": 5663
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6567215107829254,
+      "learning_rate": 7.394454508476397e-06,
+      "loss": 0.5479,
+      "step": 5664
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5775030708547746,
+      "learning_rate": 7.393548854368014e-06,
+      "loss": 0.4355,
+      "step": 5665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7233305176706283,
+      "learning_rate": 7.392643098373904e-06,
+      "loss": 0.5721,
+      "step": 5666
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1049064980849475,
+      "learning_rate": 7.391737240532624e-06,
+      "loss": 0.6325,
+      "step": 5667
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2527277130532108,
+      "learning_rate": 7.390831280882732e-06,
+      "loss": 0.6771,
+      "step": 5668
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8047751973395623,
+      "learning_rate": 7.389925219462792e-06,
+      "loss": 0.5585,
+      "step": 5669
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8064024223119045,
+      "learning_rate": 7.389019056311371e-06,
+      "loss": 0.6413,
+      "step": 5670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.589140462294574,
+      "learning_rate": 7.3881127914670455e-06,
+      "loss": 0.5318,
+      "step": 5671
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6241907795546442,
+      "learning_rate": 7.38720642496839e-06,
+      "loss": 0.5404,
+      "step": 5672
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6073188246300422,
+      "learning_rate": 7.386299956853984e-06,
+      "loss": 0.5196,
+      "step": 5673
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5545529651076304,
+      "learning_rate": 7.385393387162415e-06,
+      "loss": 0.5349,
+      "step": 5674
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7124743450162648,
+      "learning_rate": 7.384486715932272e-06,
+      "loss": 0.5394,
+      "step": 5675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5928389882833025,
+      "learning_rate": 7.383579943202148e-06,
+      "loss": 0.5678,
+      "step": 5676
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7692055633192265,
+      "learning_rate": 7.382673069010644e-06,
+      "loss": 0.5883,
+      "step": 5677
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7938091410107768,
+      "learning_rate": 7.381766093396362e-06,
+      "loss": 0.6544,
+      "step": 5678
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6892771854821916,
+      "learning_rate": 7.380859016397908e-06,
+      "loss": 0.521,
+      "step": 5679
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5692625926593775,
+      "learning_rate": 7.379951838053894e-06,
+      "loss": 0.5834,
+      "step": 5680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6058655119416744,
+      "learning_rate": 7.3790445584029345e-06,
+      "loss": 0.5688,
+      "step": 5681
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5780249326111432,
+      "learning_rate": 7.378137177483651e-06,
+      "loss": 0.5268,
+      "step": 5682
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6595996301171776,
+      "learning_rate": 7.377229695334667e-06,
+      "loss": 0.5472,
+      "step": 5683
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8328354287000123,
+      "learning_rate": 7.376322111994612e-06,
+      "loss": 0.6481,
+      "step": 5684
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.606581976116039,
+      "learning_rate": 7.375414427502117e-06,
+      "loss": 0.5033,
+      "step": 5685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.12974718942241,
+      "learning_rate": 7.374506641895822e-06,
+      "loss": 0.5692,
+      "step": 5686
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8017090667085868,
+      "learning_rate": 7.373598755214367e-06,
+      "loss": 0.5877,
+      "step": 5687
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8565991649038176,
+      "learning_rate": 7.3726907674963975e-06,
+      "loss": 0.6385,
+      "step": 5688
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9003083774224998,
+      "learning_rate": 7.371782678780563e-06,
+      "loss": 0.6021,
+      "step": 5689
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8611439299483281,
+      "learning_rate": 7.370874489105521e-06,
+      "loss": 0.6344,
+      "step": 5690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7747739841680015,
+      "learning_rate": 7.369966198509927e-06,
+      "loss": 0.6773,
+      "step": 5691
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5755643231676336,
+      "learning_rate": 7.369057807032446e-06,
+      "loss": 0.5079,
+      "step": 5692
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5969373227799505,
+      "learning_rate": 7.368149314711745e-06,
+      "loss": 0.5407,
+      "step": 5693
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6443221645439203,
+      "learning_rate": 7.367240721586493e-06,
+      "loss": 0.609,
+      "step": 5694
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8416427703897001,
+      "learning_rate": 7.3663320276953695e-06,
+      "loss": 0.6508,
+      "step": 5695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.641847549599186,
+      "learning_rate": 7.3654232330770535e-06,
+      "loss": 0.6177,
+      "step": 5696
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.582604822902364,
+      "learning_rate": 7.3645143377702284e-06,
+      "loss": 0.5219,
+      "step": 5697
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8213505261178368,
+      "learning_rate": 7.363605341813585e-06,
+      "loss": 0.5864,
+      "step": 5698
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6305030534520216,
+      "learning_rate": 7.362696245245815e-06,
+      "loss": 0.5217,
+      "step": 5699
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6773320674301949,
+      "learning_rate": 7.3617870481056165e-06,
+      "loss": 0.5949,
+      "step": 5700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6919005551557165,
+      "learning_rate": 7.360877750431689e-06,
+      "loss": 0.4858,
+      "step": 5701
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5804239232357055,
+      "learning_rate": 7.3599683522627405e-06,
+      "loss": 0.5496,
+      "step": 5702
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6610032405258903,
+      "learning_rate": 7.3590588536374805e-06,
+      "loss": 0.4988,
+      "step": 5703
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8855582237760503,
+      "learning_rate": 7.358149254594624e-06,
+      "loss": 0.596,
+      "step": 5704
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5636212708018711,
+      "learning_rate": 7.35723955517289e-06,
+      "loss": 0.4766,
+      "step": 5705
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6337044899930441,
+      "learning_rate": 7.3563297554110005e-06,
+      "loss": 0.5057,
+      "step": 5706
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6345555559804692,
+      "learning_rate": 7.35541985534768e-06,
+      "loss": 0.5733,
+      "step": 5707
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7950576407901063,
+      "learning_rate": 7.354509855021667e-06,
+      "loss": 0.555,
+      "step": 5708
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6605766481544945,
+      "learning_rate": 7.353599754471692e-06,
+      "loss": 0.5253,
+      "step": 5709
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7430219107069737,
+      "learning_rate": 7.3526895537364965e-06,
+      "loss": 0.6322,
+      "step": 5710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8479109464219405,
+      "learning_rate": 7.351779252854825e-06,
+      "loss": 0.6321,
+      "step": 5711
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.677630186932484,
+      "learning_rate": 7.350868851865426e-06,
+      "loss": 0.5457,
+      "step": 5712
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9251062126657649,
+      "learning_rate": 7.349958350807052e-06,
+      "loss": 0.5952,
+      "step": 5713
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5326632761752162,
+      "learning_rate": 7.34904774971846e-06,
+      "loss": 0.5065,
+      "step": 5714
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6292827983477869,
+      "learning_rate": 7.348137048638412e-06,
+      "loss": 0.492,
+      "step": 5715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5743666733056835,
+      "learning_rate": 7.347226247605673e-06,
+      "loss": 0.5412,
+      "step": 5716
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6139657094352567,
+      "learning_rate": 7.346315346659014e-06,
+      "loss": 0.514,
+      "step": 5717
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9149782753433823,
+      "learning_rate": 7.345404345837209e-06,
+      "loss": 0.5899,
+      "step": 5718
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7233615211723191,
+      "learning_rate": 7.344493245179035e-06,
+      "loss": 0.6705,
+      "step": 5719
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9639616574622281,
+      "learning_rate": 7.343582044723276e-06,
+      "loss": 0.5781,
+      "step": 5720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5957331894293145,
+      "learning_rate": 7.3426707445087184e-06,
+      "loss": 0.4817,
+      "step": 5721
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7178498297909918,
+      "learning_rate": 7.341759344574152e-06,
+      "loss": 0.5127,
+      "step": 5722
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8664327293580654,
+      "learning_rate": 7.340847844958374e-06,
+      "loss": 0.6246,
+      "step": 5723
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8089286751783417,
+      "learning_rate": 7.339936245700185e-06,
+      "loss": 0.6295,
+      "step": 5724
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7828796500291207,
+      "learning_rate": 7.339024546838387e-06,
+      "loss": 0.6233,
+      "step": 5725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7008235148702397,
+      "learning_rate": 7.338112748411788e-06,
+      "loss": 0.5881,
+      "step": 5726
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8967780490066428,
+      "learning_rate": 7.3372008504592004e-06,
+      "loss": 0.6183,
+      "step": 5727
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8810109259499682,
+      "learning_rate": 7.3362888530194424e-06,
+      "loss": 0.6314,
+      "step": 5728
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6175716689203922,
+      "learning_rate": 7.335376756131332e-06,
+      "loss": 0.4422,
+      "step": 5729
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7423435893389143,
+      "learning_rate": 7.334464559833696e-06,
+      "loss": 0.6334,
+      "step": 5730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6636192297928335,
+      "learning_rate": 7.3335522641653646e-06,
+      "loss": 0.5319,
+      "step": 5731
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.66616175987784,
+      "learning_rate": 7.33263986916517e-06,
+      "loss": 0.5524,
+      "step": 5732
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6475477908293895,
+      "learning_rate": 7.33172737487195e-06,
+      "loss": 0.5317,
+      "step": 5733
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7146391529957522,
+      "learning_rate": 7.330814781324547e-06,
+      "loss": 0.5766,
+      "step": 5734
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7788520304420267,
+      "learning_rate": 7.329902088561806e-06,
+      "loss": 0.6296,
+      "step": 5735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6937233753639961,
+      "learning_rate": 7.328989296622581e-06,
+      "loss": 0.6616,
+      "step": 5736
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8307538059154279,
+      "learning_rate": 7.328076405545722e-06,
+      "loss": 0.5748,
+      "step": 5737
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9242323931259457,
+      "learning_rate": 7.327163415370089e-06,
+      "loss": 0.6153,
+      "step": 5738
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9070685116030577,
+      "learning_rate": 7.326250326134548e-06,
+      "loss": 0.6259,
+      "step": 5739
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7604768170118185,
+      "learning_rate": 7.325337137877963e-06,
+      "loss": 0.6053,
+      "step": 5740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7182055685307104,
+      "learning_rate": 7.324423850639207e-06,
+      "loss": 0.536,
+      "step": 5741
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6482156691808076,
+      "learning_rate": 7.323510464457156e-06,
+      "loss": 0.4842,
+      "step": 5742
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7466382047438345,
+      "learning_rate": 7.322596979370689e-06,
+      "loss": 0.628,
+      "step": 5743
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6334318570349889,
+      "learning_rate": 7.321683395418691e-06,
+      "loss": 0.541,
+      "step": 5744
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5639089463078808,
+      "learning_rate": 7.320769712640048e-06,
+      "loss": 0.5475,
+      "step": 5745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0410641630002657,
+      "learning_rate": 7.319855931073656e-06,
+      "loss": 0.6268,
+      "step": 5746
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9442298973443018,
+      "learning_rate": 7.318942050758411e-06,
+      "loss": 0.6891,
+      "step": 5747
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6236351149290799,
+      "learning_rate": 7.318028071733212e-06,
+      "loss": 0.5682,
+      "step": 5748
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6115715915318616,
+      "learning_rate": 7.317113994036967e-06,
+      "loss": 0.5258,
+      "step": 5749
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5883876225640409,
+      "learning_rate": 7.316199817708581e-06,
+      "loss": 0.5212,
+      "step": 5750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5565977068422248,
+      "learning_rate": 7.315285542786972e-06,
+      "loss": 0.4633,
+      "step": 5751
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5742294527078042,
+      "learning_rate": 7.314371169311056e-06,
+      "loss": 0.4993,
+      "step": 5752
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7491745457961114,
+      "learning_rate": 7.313456697319754e-06,
+      "loss": 0.6878,
+      "step": 5753
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6489124135714357,
+      "learning_rate": 7.312542126851994e-06,
+      "loss": 0.5235,
+      "step": 5754
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.662751538576484,
+      "learning_rate": 7.311627457946705e-06,
+      "loss": 0.5554,
+      "step": 5755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7297963370112601,
+      "learning_rate": 7.310712690642823e-06,
+      "loss": 0.5713,
+      "step": 5756
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6276117480430634,
+      "learning_rate": 7.309797824979283e-06,
+      "loss": 0.6388,
+      "step": 5757
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8929000216587889,
+      "learning_rate": 7.308882860995034e-06,
+      "loss": 0.6366,
+      "step": 5758
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5414804000899441,
+      "learning_rate": 7.307967798729019e-06,
+      "loss": 0.4763,
+      "step": 5759
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6689040892470147,
+      "learning_rate": 7.307052638220189e-06,
+      "loss": 0.5,
+      "step": 5760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6150035201513734,
+      "learning_rate": 7.306137379507501e-06,
+      "loss": 0.5282,
+      "step": 5761
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6229377456891717,
+      "learning_rate": 7.305222022629914e-06,
+      "loss": 0.6035,
+      "step": 5762
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5776227052583409,
+      "learning_rate": 7.304306567626391e-06,
+      "loss": 0.5294,
+      "step": 5763
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.58151431386689,
+      "learning_rate": 7.3033910145359025e-06,
+      "loss": 0.5283,
+      "step": 5764
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6479035031175571,
+      "learning_rate": 7.302475363397419e-06,
+      "loss": 0.518,
+      "step": 5765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7282825080334686,
+      "learning_rate": 7.301559614249915e-06,
+      "loss": 0.5657,
+      "step": 5766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.574330821080527,
+      "learning_rate": 7.300643767132376e-06,
+      "loss": 0.5208,
+      "step": 5767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.736684147857944,
+      "learning_rate": 7.299727822083782e-06,
+      "loss": 0.6046,
+      "step": 5768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.891767415071248,
+      "learning_rate": 7.298811779143122e-06,
+      "loss": 0.618,
+      "step": 5769
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6490241774211478,
+      "learning_rate": 7.297895638349392e-06,
+      "loss": 0.5211,
+      "step": 5770
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6065242278724827,
+      "learning_rate": 7.2969793997415885e-06,
+      "loss": 0.5445,
+      "step": 5771
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5399168870417422,
+      "learning_rate": 7.2960630633587115e-06,
+      "loss": 0.503,
+      "step": 5772
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7694682650140962,
+      "learning_rate": 7.295146629239767e-06,
+      "loss": 0.5799,
+      "step": 5773
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.748506408540628,
+      "learning_rate": 7.2942300974237644e-06,
+      "loss": 0.609,
+      "step": 5774
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7446645910047759,
+      "learning_rate": 7.293313467949719e-06,
+      "loss": 0.5374,
+      "step": 5775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.707044546472477,
+      "learning_rate": 7.292396740856645e-06,
+      "loss": 0.5104,
+      "step": 5776
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5935709965582329,
+      "learning_rate": 7.291479916183571e-06,
+      "loss": 0.5092,
+      "step": 5777
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6715587002930182,
+      "learning_rate": 7.2905629939695165e-06,
+      "loss": 0.5302,
+      "step": 5778
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7167607621852933,
+      "learning_rate": 7.289645974253517e-06,
+      "loss": 0.5911,
+      "step": 5779
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5748019431123031,
+      "learning_rate": 7.288728857074603e-06,
+      "loss": 0.552,
+      "step": 5780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5809459574804884,
+      "learning_rate": 7.2878116424718156e-06,
+      "loss": 0.5267,
+      "step": 5781
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7847818709982552,
+      "learning_rate": 7.286894330484199e-06,
+      "loss": 0.6018,
+      "step": 5782
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7157408343451143,
+      "learning_rate": 7.285976921150797e-06,
+      "loss": 0.5727,
+      "step": 5783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.597567504021634,
+      "learning_rate": 7.285059414510662e-06,
+      "loss": 0.5119,
+      "step": 5784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5660599563764835,
+      "learning_rate": 7.284141810602851e-06,
+      "loss": 0.4974,
+      "step": 5785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7149585941121873,
+      "learning_rate": 7.283224109466422e-06,
+      "loss": 0.5979,
+      "step": 5786
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7330929300573952,
+      "learning_rate": 7.282306311140439e-06,
+      "loss": 0.6008,
+      "step": 5787
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5428974476513099,
+      "learning_rate": 7.281388415663969e-06,
+      "loss": 0.5337,
+      "step": 5788
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5691331466082877,
+      "learning_rate": 7.280470423076085e-06,
+      "loss": 0.5148,
+      "step": 5789
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6695316733450618,
+      "learning_rate": 7.279552333415862e-06,
+      "loss": 0.5428,
+      "step": 5790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7811793793546558,
+      "learning_rate": 7.278634146722381e-06,
+      "loss": 0.5439,
+      "step": 5791
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7569573819473664,
+      "learning_rate": 7.277715863034725e-06,
+      "loss": 0.6132,
+      "step": 5792
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6489098647452222,
+      "learning_rate": 7.276797482391985e-06,
+      "loss": 0.5182,
+      "step": 5793
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8661039160399061,
+      "learning_rate": 7.27587900483325e-06,
+      "loss": 0.6641,
+      "step": 5794
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5820186757358307,
+      "learning_rate": 7.274960430397618e-06,
+      "loss": 0.5221,
+      "step": 5795
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6325369896018755,
+      "learning_rate": 7.274041759124192e-06,
+      "loss": 0.5318,
+      "step": 5796
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7142674492910979,
+      "learning_rate": 7.273122991052074e-06,
+      "loss": 0.6028,
+      "step": 5797
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5693571249743329,
+      "learning_rate": 7.272204126220375e-06,
+      "loss": 0.5323,
+      "step": 5798
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5726341623341924,
+      "learning_rate": 7.271285164668207e-06,
+      "loss": 0.4993,
+      "step": 5799
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5720652830387938,
+      "learning_rate": 7.270366106434687e-06,
+      "loss": 0.5106,
+      "step": 5800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5975392123756584,
+      "learning_rate": 7.269446951558936e-06,
+      "loss": 0.5266,
+      "step": 5801
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6988535752173115,
+      "learning_rate": 7.268527700080081e-06,
+      "loss": 0.5369,
+      "step": 5802
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6046355646140684,
+      "learning_rate": 7.267608352037252e-06,
+      "loss": 0.4992,
+      "step": 5803
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6186768633041764,
+      "learning_rate": 7.266688907469582e-06,
+      "loss": 0.5125,
+      "step": 5804
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7500248355786981,
+      "learning_rate": 7.2657693664162074e-06,
+      "loss": 0.5671,
+      "step": 5805
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7928151669816452,
+      "learning_rate": 7.264849728916272e-06,
+      "loss": 0.6154,
+      "step": 5806
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0398998040134704,
+      "learning_rate": 7.263929995008921e-06,
+      "loss": 0.6265,
+      "step": 5807
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6269689146676133,
+      "learning_rate": 7.263010164733303e-06,
+      "loss": 0.4923,
+      "step": 5808
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7283140190877309,
+      "learning_rate": 7.262090238128575e-06,
+      "loss": 0.5979,
+      "step": 5809
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.836802862033315,
+      "learning_rate": 7.261170215233895e-06,
+      "loss": 0.6063,
+      "step": 5810
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8295010847206123,
+      "learning_rate": 7.2602500960884235e-06,
+      "loss": 0.6066,
+      "step": 5811
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7460122452114994,
+      "learning_rate": 7.259329880731328e-06,
+      "loss": 0.6002,
+      "step": 5812
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5560955965994837,
+      "learning_rate": 7.2584095692017795e-06,
+      "loss": 0.5122,
+      "step": 5813
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7079701401655037,
+      "learning_rate": 7.257489161538953e-06,
+      "loss": 0.5133,
+      "step": 5814
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.142037142805699,
+      "learning_rate": 7.256568657782026e-06,
+      "loss": 0.5646,
+      "step": 5815
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6196514172605062,
+      "learning_rate": 7.255648057970184e-06,
+      "loss": 0.5278,
+      "step": 5816
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.660998859609431,
+      "learning_rate": 7.254727362142611e-06,
+      "loss": 0.5407,
+      "step": 5817
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7939894477975382,
+      "learning_rate": 7.253806570338499e-06,
+      "loss": 0.6261,
+      "step": 5818
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7881642132198244,
+      "learning_rate": 7.252885682597041e-06,
+      "loss": 0.5966,
+      "step": 5819
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5466048211434927,
+      "learning_rate": 7.2519646989574435e-06,
+      "loss": 0.5176,
+      "step": 5820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8007518272564078,
+      "learning_rate": 7.251043619458902e-06,
+      "loss": 0.5782,
+      "step": 5821
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7113264111127049,
+      "learning_rate": 7.250122444140628e-06,
+      "loss": 0.6143,
+      "step": 5822
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5343792154273836,
+      "learning_rate": 7.249201173041832e-06,
+      "loss": 0.4949,
+      "step": 5823
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6462591485954078,
+      "learning_rate": 7.248279806201729e-06,
+      "loss": 0.5439,
+      "step": 5824
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.70629652786406,
+      "learning_rate": 7.2473583436595384e-06,
+      "loss": 0.5427,
+      "step": 5825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.633347607682042,
+      "learning_rate": 7.246436785454486e-06,
+      "loss": 0.5787,
+      "step": 5826
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7389954359806247,
+      "learning_rate": 7.2455151316257975e-06,
+      "loss": 0.5875,
+      "step": 5827
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8314776720107688,
+      "learning_rate": 7.244593382212706e-06,
+      "loss": 0.5708,
+      "step": 5828
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5866210514358545,
+      "learning_rate": 7.243671537254446e-06,
+      "loss": 0.5067,
+      "step": 5829
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6535199784271271,
+      "learning_rate": 7.24274959679026e-06,
+      "loss": 0.5729,
+      "step": 5830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7643511709475582,
+      "learning_rate": 7.241827560859388e-06,
+      "loss": 0.5796,
+      "step": 5831
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8033218630139487,
+      "learning_rate": 7.240905429501083e-06,
+      "loss": 0.6167,
+      "step": 5832
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.264851543870776,
+      "learning_rate": 7.239983202754594e-06,
+      "loss": 0.5799,
+      "step": 5833
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6654171955223167,
+      "learning_rate": 7.239060880659177e-06,
+      "loss": 0.5984,
+      "step": 5834
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.635724976641325,
+      "learning_rate": 7.238138463254095e-06,
+      "loss": 0.4765,
+      "step": 5835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6263943895827097,
+      "learning_rate": 7.23721595057861e-06,
+      "loss": 0.6167,
+      "step": 5836
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7289171218932294,
+      "learning_rate": 7.2362933426719905e-06,
+      "loss": 0.5762,
+      "step": 5837
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6279899523271233,
+      "learning_rate": 7.235370639573509e-06,
+      "loss": 0.5283,
+      "step": 5838
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.710489049279531,
+      "learning_rate": 7.234447841322445e-06,
+      "loss": 0.517,
+      "step": 5839
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6623958418177638,
+      "learning_rate": 7.233524947958075e-06,
+      "loss": 0.5515,
+      "step": 5840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6971840183623577,
+      "learning_rate": 7.232601959519685e-06,
+      "loss": 0.5163,
+      "step": 5841
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.631580799135298,
+      "learning_rate": 7.231678876046565e-06,
+      "loss": 0.5337,
+      "step": 5842
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8002069483257911,
+      "learning_rate": 7.230755697578007e-06,
+      "loss": 0.6165,
+      "step": 5843
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8090719609189326,
+      "learning_rate": 7.229832424153307e-06,
+      "loss": 0.6482,
+      "step": 5844
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6438418091126863,
+      "learning_rate": 7.228909055811766e-06,
+      "loss": 0.547,
+      "step": 5845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6686358657763787,
+      "learning_rate": 7.227985592592688e-06,
+      "loss": 0.606,
+      "step": 5846
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5883747889183144,
+      "learning_rate": 7.227062034535384e-06,
+      "loss": 0.5372,
+      "step": 5847
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7887012703417479,
+      "learning_rate": 7.226138381679165e-06,
+      "loss": 0.6289,
+      "step": 5848
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8359052276538436,
+      "learning_rate": 7.22521463406335e-06,
+      "loss": 0.6622,
+      "step": 5849
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8624524633674899,
+      "learning_rate": 7.224290791727259e-06,
+      "loss": 0.6222,
+      "step": 5850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6247450915222917,
+      "learning_rate": 7.223366854710216e-06,
+      "loss": 0.5271,
+      "step": 5851
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7475055977408548,
+      "learning_rate": 7.222442823051552e-06,
+      "loss": 0.6031,
+      "step": 5852
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7956688269029737,
+      "learning_rate": 7.221518696790597e-06,
+      "loss": 0.621,
+      "step": 5853
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5937287743492685,
+      "learning_rate": 7.220594475966691e-06,
+      "loss": 0.5155,
+      "step": 5854
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5664963442322977,
+      "learning_rate": 7.219670160619174e-06,
+      "loss": 0.5153,
+      "step": 5855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9856919355818669,
+      "learning_rate": 7.218745750787392e-06,
+      "loss": 0.6404,
+      "step": 5856
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6901444752687154,
+      "learning_rate": 7.217821246510692e-06,
+      "loss": 0.565,
+      "step": 5857
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7678632427900924,
+      "learning_rate": 7.216896647828431e-06,
+      "loss": 0.6458,
+      "step": 5858
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6206510986093188,
+      "learning_rate": 7.215971954779962e-06,
+      "loss": 0.5602,
+      "step": 5859
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7804973788476395,
+      "learning_rate": 7.21504716740465e-06,
+      "loss": 0.6281,
+      "step": 5860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7043479877195835,
+      "learning_rate": 7.2141222857418555e-06,
+      "loss": 0.5742,
+      "step": 5861
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7256562511537287,
+      "learning_rate": 7.213197309830953e-06,
+      "loss": 0.5702,
+      "step": 5862
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6173465471508519,
+      "learning_rate": 7.212272239711312e-06,
+      "loss": 0.5248,
+      "step": 5863
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6334666783535389,
+      "learning_rate": 7.211347075422312e-06,
+      "loss": 0.5136,
+      "step": 5864
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7329613183016386,
+      "learning_rate": 7.2104218170033325e-06,
+      "loss": 0.5893,
+      "step": 5865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8142819818199812,
+      "learning_rate": 7.209496464493762e-06,
+      "loss": 0.6028,
+      "step": 5866
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6844891015001855,
+      "learning_rate": 7.208571017932985e-06,
+      "loss": 0.5128,
+      "step": 5867
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8260304063851073,
+      "learning_rate": 7.207645477360399e-06,
+      "loss": 0.5968,
+      "step": 5868
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6660889071416392,
+      "learning_rate": 7.206719842815398e-06,
+      "loss": 0.5666,
+      "step": 5869
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9208399051921774,
+      "learning_rate": 7.205794114337385e-06,
+      "loss": 0.6336,
+      "step": 5870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6119512037323821,
+      "learning_rate": 7.204868291965767e-06,
+      "loss": 0.544,
+      "step": 5871
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6525157575923063,
+      "learning_rate": 7.203942375739951e-06,
+      "loss": 0.575,
+      "step": 5872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6777254365784581,
+      "learning_rate": 7.2030163656993505e-06,
+      "loss": 0.5898,
+      "step": 5873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5781114453839473,
+      "learning_rate": 7.2020902618833836e-06,
+      "loss": 0.535,
+      "step": 5874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7283837867809089,
+      "learning_rate": 7.201164064331469e-06,
+      "loss": 0.6501,
+      "step": 5875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6009244489423754,
+      "learning_rate": 7.200237773083036e-06,
+      "loss": 0.5303,
+      "step": 5876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7275894205842025,
+      "learning_rate": 7.199311388177512e-06,
+      "loss": 0.4724,
+      "step": 5877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.643228658523126,
+      "learning_rate": 7.198384909654331e-06,
+      "loss": 0.5359,
+      "step": 5878
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7513516441067521,
+      "learning_rate": 7.197458337552928e-06,
+      "loss": 0.605,
+      "step": 5879
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6258683148799139,
+      "learning_rate": 7.1965316719127465e-06,
+      "loss": 0.5499,
+      "step": 5880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7772835189884585,
+      "learning_rate": 7.1956049127732295e-06,
+      "loss": 0.5744,
+      "step": 5881
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6216236494210087,
+      "learning_rate": 7.1946780601738285e-06,
+      "loss": 0.5491,
+      "step": 5882
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5663351592564475,
+      "learning_rate": 7.1937511141539965e-06,
+      "loss": 0.5244,
+      "step": 5883
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7554566424136938,
+      "learning_rate": 7.192824074753188e-06,
+      "loss": 0.5798,
+      "step": 5884
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7613029883601454,
+      "learning_rate": 7.191896942010867e-06,
+      "loss": 0.599,
+      "step": 5885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8065786710297403,
+      "learning_rate": 7.190969715966498e-06,
+      "loss": 0.6365,
+      "step": 5886
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5736921731008394,
+      "learning_rate": 7.190042396659548e-06,
+      "loss": 0.5074,
+      "step": 5887
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6962951030751786,
+      "learning_rate": 7.189114984129492e-06,
+      "loss": 0.6247,
+      "step": 5888
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6855375517979899,
+      "learning_rate": 7.1881874784158065e-06,
+      "loss": 0.5474,
+      "step": 5889
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7174166678113488,
+      "learning_rate": 7.187259879557974e-06,
+      "loss": 0.5967,
+      "step": 5890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.746697031944723,
+      "learning_rate": 7.186332187595477e-06,
+      "loss": 0.5463,
+      "step": 5891
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6994200247041723,
+      "learning_rate": 7.185404402567805e-06,
+      "loss": 0.6516,
+      "step": 5892
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8037768005668379,
+      "learning_rate": 7.18447652451445e-06,
+      "loss": 0.5931,
+      "step": 5893
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8136246375220463,
+      "learning_rate": 7.183548553474912e-06,
+      "loss": 0.5129,
+      "step": 5894
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5920517296472912,
+      "learning_rate": 7.182620489488689e-06,
+      "loss": 0.5559,
+      "step": 5895
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7856583055030771,
+      "learning_rate": 7.181692332595286e-06,
+      "loss": 0.6187,
+      "step": 5896
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8117548383096944,
+      "learning_rate": 7.180764082834213e-06,
+      "loss": 0.6325,
+      "step": 5897
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8346757934061227,
+      "learning_rate": 7.1798357402449805e-06,
+      "loss": 0.6566,
+      "step": 5898
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7666718394886585,
+      "learning_rate": 7.178907304867108e-06,
+      "loss": 0.6011,
+      "step": 5899
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6797956495444377,
+      "learning_rate": 7.177978776740112e-06,
+      "loss": 0.5266,
+      "step": 5900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5985809901293472,
+      "learning_rate": 7.17705015590352e-06,
+      "loss": 0.5502,
+      "step": 5901
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6179930592407553,
+      "learning_rate": 7.176121442396861e-06,
+      "loss": 0.5399,
+      "step": 5902
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.887166665193454,
+      "learning_rate": 7.175192636259666e-06,
+      "loss": 0.6518,
+      "step": 5903
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6767972616304102,
+      "learning_rate": 7.174263737531471e-06,
+      "loss": 0.5321,
+      "step": 5904
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7545027952937109,
+      "learning_rate": 7.173334746251815e-06,
+      "loss": 0.5955,
+      "step": 5905
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.694087376964566,
+      "learning_rate": 7.172405662460247e-06,
+      "loss": 0.6055,
+      "step": 5906
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5405844381160699,
+      "learning_rate": 7.1714764861963095e-06,
+      "loss": 0.5076,
+      "step": 5907
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6662841454229068,
+      "learning_rate": 7.170547217499557e-06,
+      "loss": 0.5525,
+      "step": 5908
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6366622953428074,
+      "learning_rate": 7.169617856409547e-06,
+      "loss": 0.5104,
+      "step": 5909
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7557653178685522,
+      "learning_rate": 7.168688402965837e-06,
+      "loss": 0.6106,
+      "step": 5910
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5698851246736489,
+      "learning_rate": 7.167758857207992e-06,
+      "loss": 0.5251,
+      "step": 5911
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6579949979231071,
+      "learning_rate": 7.16682921917558e-06,
+      "loss": 0.4939,
+      "step": 5912
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9267038564495211,
+      "learning_rate": 7.165899488908171e-06,
+      "loss": 0.6138,
+      "step": 5913
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6137144751518273,
+      "learning_rate": 7.1649696664453435e-06,
+      "loss": 0.4945,
+      "step": 5914
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6336469828178837,
+      "learning_rate": 7.164039751826675e-06,
+      "loss": 0.5368,
+      "step": 5915
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5332662473473622,
+      "learning_rate": 7.16310974509175e-06,
+      "loss": 0.5095,
+      "step": 5916
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6380657288145953,
+      "learning_rate": 7.162179646280155e-06,
+      "loss": 0.5039,
+      "step": 5917
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5217335558926929,
+      "learning_rate": 7.161249455431481e-06,
+      "loss": 0.5023,
+      "step": 5918
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7476408907405009,
+      "learning_rate": 7.160319172585325e-06,
+      "loss": 0.6326,
+      "step": 5919
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8369853122374212,
+      "learning_rate": 7.159388797781285e-06,
+      "loss": 0.6343,
+      "step": 5920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8397283935017545,
+      "learning_rate": 7.1584583310589635e-06,
+      "loss": 0.6006,
+      "step": 5921
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.686576869167361,
+      "learning_rate": 7.15752777245797e-06,
+      "loss": 0.5358,
+      "step": 5922
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8897349901289278,
+      "learning_rate": 7.156597122017913e-06,
+      "loss": 0.6119,
+      "step": 5923
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6827827637178683,
+      "learning_rate": 7.155666379778407e-06,
+      "loss": 0.5499,
+      "step": 5924
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6486559097349742,
+      "learning_rate": 7.154735545779072e-06,
+      "loss": 0.4898,
+      "step": 5925
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4057449600176155,
+      "learning_rate": 7.153804620059532e-06,
+      "loss": 0.6217,
+      "step": 5926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5506670451254673,
+      "learning_rate": 7.152873602659411e-06,
+      "loss": 0.5245,
+      "step": 5927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6546870274474008,
+      "learning_rate": 7.15194249361834e-06,
+      "loss": 0.5798,
+      "step": 5928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7520327982372209,
+      "learning_rate": 7.1510112929759544e-06,
+      "loss": 0.6366,
+      "step": 5929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6011870153450867,
+      "learning_rate": 7.150080000771892e-06,
+      "loss": 0.5259,
+      "step": 5930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5643861021086324,
+      "learning_rate": 7.149148617045793e-06,
+      "loss": 0.5191,
+      "step": 5931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6655891581971597,
+      "learning_rate": 7.1482171418373055e-06,
+      "loss": 0.5382,
+      "step": 5932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7097024475756677,
+      "learning_rate": 7.14728557518608e-06,
+      "loss": 0.5988,
+      "step": 5933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5814307284642501,
+      "learning_rate": 7.146353917131767e-06,
+      "loss": 0.5127,
+      "step": 5934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8005391443407877,
+      "learning_rate": 7.145422167714029e-06,
+      "loss": 0.6022,
+      "step": 5935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7009976228365044,
+      "learning_rate": 7.144490326972525e-06,
+      "loss": 0.55,
+      "step": 5936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6720608201194977,
+      "learning_rate": 7.143558394946919e-06,
+      "loss": 0.601,
+      "step": 5937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9116088358992631,
+      "learning_rate": 7.142626371676883e-06,
+      "loss": 0.6073,
+      "step": 5938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6275254875012132,
+      "learning_rate": 7.14169425720209e-06,
+      "loss": 0.527,
+      "step": 5939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5094376765807014,
+      "learning_rate": 7.140762051562215e-06,
+      "loss": 0.6293,
+      "step": 5940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6549302507765816,
+      "learning_rate": 7.139829754796941e-06,
+      "loss": 0.5013,
+      "step": 5941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.612729566792426,
+      "learning_rate": 7.1388973669459515e-06,
+      "loss": 0.5466,
+      "step": 5942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7536579417526534,
+      "learning_rate": 7.137964888048936e-06,
+      "loss": 0.6376,
+      "step": 5943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9117217692848721,
+      "learning_rate": 7.137032318145588e-06,
+      "loss": 0.6506,
+      "step": 5944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7630075700909473,
+      "learning_rate": 7.1360996572756034e-06,
+      "loss": 0.6294,
+      "step": 5945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6414206866660678,
+      "learning_rate": 7.1351669054786825e-06,
+      "loss": 0.552,
+      "step": 5946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7169280605097303,
+      "learning_rate": 7.134234062794528e-06,
+      "loss": 0.6154,
+      "step": 5947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5957026845046108,
+      "learning_rate": 7.1333011292628506e-06,
+      "loss": 0.5875,
+      "step": 5948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6799298689048163,
+      "learning_rate": 7.132368104923359e-06,
+      "loss": 0.5323,
+      "step": 5949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6339070219489517,
+      "learning_rate": 7.131434989815774e-06,
+      "loss": 0.5409,
+      "step": 5950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8221592820088667,
+      "learning_rate": 7.130501783979813e-06,
+      "loss": 0.6135,
+      "step": 5951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7197916486344914,
+      "learning_rate": 7.129568487455197e-06,
+      "loss": 0.5424,
+      "step": 5952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7347070262150646,
+      "learning_rate": 7.128635100281657e-06,
+      "loss": 0.5833,
+      "step": 5953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7357879577306148,
+      "learning_rate": 7.1277016224989224e-06,
+      "loss": 0.5877,
+      "step": 5954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8899153229202661,
+      "learning_rate": 7.126768054146729e-06,
+      "loss": 0.621,
+      "step": 5955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8080942458176608,
+      "learning_rate": 7.125834395264815e-06,
+      "loss": 0.6003,
+      "step": 5956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7409659378027762,
+      "learning_rate": 7.124900645892925e-06,
+      "loss": 0.5787,
+      "step": 5957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5636619965707601,
+      "learning_rate": 7.123966806070804e-06,
+      "loss": 0.5148,
+      "step": 5958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6939137855049864,
+      "learning_rate": 7.123032875838204e-06,
+      "loss": 0.5317,
+      "step": 5959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6169178512654427,
+      "learning_rate": 7.122098855234879e-06,
+      "loss": 0.5068,
+      "step": 5960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7670689037036561,
+      "learning_rate": 7.121164744300586e-06,
+      "loss": 0.6026,
+      "step": 5961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5868224830446753,
+      "learning_rate": 7.120230543075088e-06,
+      "loss": 0.5358,
+      "step": 5962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5930265037171247,
+      "learning_rate": 7.119296251598152e-06,
+      "loss": 0.566,
+      "step": 5963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5908456018357738,
+      "learning_rate": 7.118361869909547e-06,
+      "loss": 0.5208,
+      "step": 5964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5593592665307044,
+      "learning_rate": 7.117427398049045e-06,
+      "loss": 0.5258,
+      "step": 5965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.65059181906227,
+      "learning_rate": 7.116492836056427e-06,
+      "loss": 0.5709,
+      "step": 5966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7629845456020955,
+      "learning_rate": 7.11555818397147e-06,
+      "loss": 0.5841,
+      "step": 5967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7274199972082073,
+      "learning_rate": 7.1146234418339635e-06,
+      "loss": 0.6031,
+      "step": 5968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5852919273140857,
+      "learning_rate": 7.1136886096836935e-06,
+      "loss": 0.5653,
+      "step": 5969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7087866747568418,
+      "learning_rate": 7.112753687560454e-06,
+      "loss": 0.5917,
+      "step": 5970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.56202775250031,
+      "learning_rate": 7.111818675504041e-06,
+      "loss": 0.5322,
+      "step": 5971
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5920179015436181,
+      "learning_rate": 7.1108835735542556e-06,
+      "loss": 0.5292,
+      "step": 5972
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7633463885951367,
+      "learning_rate": 7.109948381750902e-06,
+      "loss": 0.5593,
+      "step": 5973
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6123231443472408,
+      "learning_rate": 7.109013100133789e-06,
+      "loss": 0.5001,
+      "step": 5974
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8101548896370763,
+      "learning_rate": 7.108077728742727e-06,
+      "loss": 0.6194,
+      "step": 5975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6193684625645861,
+      "learning_rate": 7.107142267617533e-06,
+      "loss": 0.5115,
+      "step": 5976
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5573063001403132,
+      "learning_rate": 7.106206716798025e-06,
+      "loss": 0.5082,
+      "step": 5977
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7036027776981749,
+      "learning_rate": 7.105271076324028e-06,
+      "loss": 0.5874,
+      "step": 5978
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.652049101670856,
+      "learning_rate": 7.10433534623537e-06,
+      "loss": 0.5015,
+      "step": 5979
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5763090882467994,
+      "learning_rate": 7.103399526571879e-06,
+      "loss": 0.5307,
+      "step": 5980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.674680021102911,
+      "learning_rate": 7.102463617373392e-06,
+      "loss": 0.5083,
+      "step": 5981
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6330173496203505,
+      "learning_rate": 7.101527618679749e-06,
+      "loss": 0.5371,
+      "step": 5982
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7500296472281367,
+      "learning_rate": 7.10059153053079e-06,
+      "loss": 0.4961,
+      "step": 5983
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7230364666967767,
+      "learning_rate": 7.099655352966361e-06,
+      "loss": 0.565,
+      "step": 5984
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7281334536321968,
+      "learning_rate": 7.098719086026313e-06,
+      "loss": 0.5901,
+      "step": 5985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5669714110538367,
+      "learning_rate": 7.097782729750501e-06,
+      "loss": 0.552,
+      "step": 5986
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5776175088594866,
+      "learning_rate": 7.0968462841787825e-06,
+      "loss": 0.5161,
+      "step": 5987
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8035701307743665,
+      "learning_rate": 7.095909749351016e-06,
+      "loss": 0.5945,
+      "step": 5988
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5749391162769852,
+      "learning_rate": 7.094973125307071e-06,
+      "loss": 0.5199,
+      "step": 5989
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6034501105002962,
+      "learning_rate": 7.094036412086815e-06,
+      "loss": 0.5526,
+      "step": 5990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6351080516770963,
+      "learning_rate": 7.093099609730118e-06,
+      "loss": 0.584,
+      "step": 5991
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.616218268716067,
+      "learning_rate": 7.092162718276861e-06,
+      "loss": 0.5596,
+      "step": 5992
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6133377103476508,
+      "learning_rate": 7.091225737766923e-06,
+      "loss": 0.549,
+      "step": 5993
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6974070658555137,
+      "learning_rate": 7.090288668240187e-06,
+      "loss": 0.5781,
+      "step": 5994
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6675311153928515,
+      "learning_rate": 7.089351509736543e-06,
+      "loss": 0.5763,
+      "step": 5995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.756118850204332,
+      "learning_rate": 7.0884142622958805e-06,
+      "loss": 0.6154,
+      "step": 5996
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6698436678644841,
+      "learning_rate": 7.087476925958098e-06,
+      "loss": 0.5872,
+      "step": 5997
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5878670451744555,
+      "learning_rate": 7.086539500763092e-06,
+      "loss": 0.5009,
+      "step": 5998
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6414485430906762,
+      "learning_rate": 7.085601986750767e-06,
+      "loss": 0.5466,
+      "step": 5999
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7238108126282299,
+      "learning_rate": 7.084664383961031e-06,
+      "loss": 0.5699,
+      "step": 6000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6678033522689986,
+      "learning_rate": 7.083726692433793e-06,
+      "loss": 0.6002,
+      "step": 6001
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7837549105544315,
+      "learning_rate": 7.08278891220897e-06,
+      "loss": 0.6126,
+      "step": 6002
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6074108868955742,
+      "learning_rate": 7.081851043326477e-06,
+      "loss": 0.5207,
+      "step": 6003
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7707852242079446,
+      "learning_rate": 7.0809130858262385e-06,
+      "loss": 0.5875,
+      "step": 6004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8314941488122427,
+      "learning_rate": 7.079975039748179e-06,
+      "loss": 0.681,
+      "step": 6005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7864963641232625,
+      "learning_rate": 7.0790369051322285e-06,
+      "loss": 0.6382,
+      "step": 6006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.545907604352695,
+      "learning_rate": 7.0780986820183215e-06,
+      "loss": 0.5221,
+      "step": 6007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8486654749552658,
+      "learning_rate": 7.077160370446395e-06,
+      "loss": 0.6261,
+      "step": 6008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8072855106678415,
+      "learning_rate": 7.076221970456387e-06,
+      "loss": 0.6484,
+      "step": 6009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5287191367170836,
+      "learning_rate": 7.075283482088246e-06,
+      "loss": 0.4511,
+      "step": 6010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5741171173985696,
+      "learning_rate": 7.074344905381917e-06,
+      "loss": 0.5015,
+      "step": 6011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1406092082545656,
+      "learning_rate": 7.073406240377356e-06,
+      "loss": 0.5978,
+      "step": 6012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7210962449068865,
+      "learning_rate": 7.072467487114516e-06,
+      "loss": 0.5889,
+      "step": 6013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9730920088064027,
+      "learning_rate": 7.071528645633358e-06,
+      "loss": 0.6462,
+      "step": 6014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5686619356341395,
+      "learning_rate": 7.070589715973845e-06,
+      "loss": 0.5222,
+      "step": 6015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.582273852585019,
+      "learning_rate": 7.069650698175945e-06,
+      "loss": 0.5512,
+      "step": 6016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5794857780835623,
+      "learning_rate": 7.068711592279628e-06,
+      "loss": 0.5046,
+      "step": 6017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8308853893111957,
+      "learning_rate": 7.067772398324869e-06,
+      "loss": 0.6769,
+      "step": 6018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8871208267240814,
+      "learning_rate": 7.0668331163516475e-06,
+      "loss": 0.6243,
+      "step": 6019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7353649761204798,
+      "learning_rate": 7.065893746399945e-06,
+      "loss": 0.6082,
+      "step": 6020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8206904928576645,
+      "learning_rate": 7.0649542885097464e-06,
+      "loss": 0.6601,
+      "step": 6021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7994782475497233,
+      "learning_rate": 7.064014742721043e-06,
+      "loss": 0.5921,
+      "step": 6022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6300920174866359,
+      "learning_rate": 7.063075109073827e-06,
+      "loss": 0.5324,
+      "step": 6023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7172389415559045,
+      "learning_rate": 7.062135387608095e-06,
+      "loss": 0.5691,
+      "step": 6024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6011265374447058,
+      "learning_rate": 7.0611955783638516e-06,
+      "loss": 0.5494,
+      "step": 6025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7913267524061012,
+      "learning_rate": 7.060255681381098e-06,
+      "loss": 0.5861,
+      "step": 6026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5408541454873202,
+      "learning_rate": 7.0593156966998445e-06,
+      "loss": 0.4974,
+      "step": 6027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7635874309264382,
+      "learning_rate": 7.058375624360102e-06,
+      "loss": 0.6004,
+      "step": 6028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5356125016899917,
+      "learning_rate": 7.057435464401887e-06,
+      "loss": 0.5235,
+      "step": 6029
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.626775555298914,
+      "learning_rate": 7.056495216865219e-06,
+      "loss": 0.5085,
+      "step": 6030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8547834497440239,
+      "learning_rate": 7.055554881790121e-06,
+      "loss": 0.6407,
+      "step": 6031
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7310096864764011,
+      "learning_rate": 7.054614459216619e-06,
+      "loss": 0.6107,
+      "step": 6032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5757760455675806,
+      "learning_rate": 7.0536739491847465e-06,
+      "loss": 0.5305,
+      "step": 6033
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7645665550905649,
+      "learning_rate": 7.052733351734537e-06,
+      "loss": 0.6176,
+      "step": 6034
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8265033598092296,
+      "learning_rate": 7.051792666906027e-06,
+      "loss": 0.6061,
+      "step": 6035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6837039457234325,
+      "learning_rate": 7.0508518947392614e-06,
+      "loss": 0.5634,
+      "step": 6036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.586998764721055,
+      "learning_rate": 7.049911035274283e-06,
+      "loss": 0.5124,
+      "step": 6037
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5892951271709775,
+      "learning_rate": 7.048970088551143e-06,
+      "loss": 0.517,
+      "step": 6038
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6512704912070886,
+      "learning_rate": 7.048029054609893e-06,
+      "loss": 0.5707,
+      "step": 6039
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7442628176487647,
+      "learning_rate": 7.047087933490591e-06,
+      "loss": 0.597,
+      "step": 6040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6634565714644403,
+      "learning_rate": 7.0461467252332984e-06,
+      "loss": 0.4795,
+      "step": 6041
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7873848421662484,
+      "learning_rate": 7.045205429878076e-06,
+      "loss": 0.6211,
+      "step": 6042
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6345169378203459,
+      "learning_rate": 7.044264047464995e-06,
+      "loss": 0.4902,
+      "step": 6043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6048493037612452,
+      "learning_rate": 7.0433225780341265e-06,
+      "loss": 0.5271,
+      "step": 6044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5987685909027206,
+      "learning_rate": 7.042381021625546e-06,
+      "loss": 0.5846,
+      "step": 6045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.761014568982348,
+      "learning_rate": 7.0414393782793315e-06,
+      "loss": 0.5534,
+      "step": 6046
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5899132072894913,
+      "learning_rate": 7.040497648035565e-06,
+      "loss": 0.5201,
+      "step": 6047
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7087852389345419,
+      "learning_rate": 7.0395558309343345e-06,
+      "loss": 0.6047,
+      "step": 6048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.877431540022819,
+      "learning_rate": 7.03861392701573e-06,
+      "loss": 0.6155,
+      "step": 6049
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5516692086699265,
+      "learning_rate": 7.037671936319845e-06,
+      "loss": 0.4954,
+      "step": 6050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5984557294666936,
+      "learning_rate": 7.036729858886779e-06,
+      "loss": 0.5467,
+      "step": 6051
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5747723406091592,
+      "learning_rate": 7.035787694756629e-06,
+      "loss": 0.547,
+      "step": 6052
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6071889743209812,
+      "learning_rate": 7.0348454439695035e-06,
+      "loss": 0.5746,
+      "step": 6053
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7304023891510193,
+      "learning_rate": 7.033903106565509e-06,
+      "loss": 0.6284,
+      "step": 6054
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6570541196986908,
+      "learning_rate": 7.032960682584758e-06,
+      "loss": 0.6438,
+      "step": 6055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5953741344451474,
+      "learning_rate": 7.03201817206737e-06,
+      "loss": 0.5526,
+      "step": 6056
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7390360241547438,
+      "learning_rate": 7.031075575053459e-06,
+      "loss": 0.6218,
+      "step": 6057
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6182872596546606,
+      "learning_rate": 7.030132891583152e-06,
+      "loss": 0.55,
+      "step": 6058
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5474958207745392,
+      "learning_rate": 7.029190121696577e-06,
+      "loss": 0.496,
+      "step": 6059
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8319296740253522,
+      "learning_rate": 7.028247265433861e-06,
+      "loss": 0.6289,
+      "step": 6060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6632324180366329,
+      "learning_rate": 7.0273043228351404e-06,
+      "loss": 0.6,
+      "step": 6061
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7093942999794963,
+      "learning_rate": 7.0263612939405524e-06,
+      "loss": 0.568,
+      "step": 6062
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.672908521751254,
+      "learning_rate": 7.025418178790241e-06,
+      "loss": 0.5213,
+      "step": 6063
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.880776464295765,
+      "learning_rate": 7.0244749774243495e-06,
+      "loss": 0.5648,
+      "step": 6064
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6353285734379588,
+      "learning_rate": 7.0235316898830274e-06,
+      "loss": 0.5592,
+      "step": 6065
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.80600462935185,
+      "learning_rate": 7.022588316206427e-06,
+      "loss": 0.664,
+      "step": 6066
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6758317174509741,
+      "learning_rate": 7.021644856434705e-06,
+      "loss": 0.5806,
+      "step": 6067
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6523254610023184,
+      "learning_rate": 7.0207013106080225e-06,
+      "loss": 0.5572,
+      "step": 6068
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5382601567959718,
+      "learning_rate": 7.019757678766544e-06,
+      "loss": 0.541,
+      "step": 6069
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6813693905307426,
+      "learning_rate": 7.018813960950432e-06,
+      "loss": 0.5508,
+      "step": 6070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.745718707449539,
+      "learning_rate": 7.0178701571998644e-06,
+      "loss": 0.5573,
+      "step": 6071
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6610284907289825,
+      "learning_rate": 7.01692626755501e-06,
+      "loss": 0.5677,
+      "step": 6072
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7525283827944745,
+      "learning_rate": 7.01598229205605e-06,
+      "loss": 0.5415,
+      "step": 6073
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6812398425879042,
+      "learning_rate": 7.015038230743165e-06,
+      "loss": 0.5766,
+      "step": 6074
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6376153279099177,
+      "learning_rate": 7.014094083656544e-06,
+      "loss": 0.5339,
+      "step": 6075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6617176435761051,
+      "learning_rate": 7.013149850836373e-06,
+      "loss": 0.5468,
+      "step": 6076
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.612352892101539,
+      "learning_rate": 7.0122055323228455e-06,
+      "loss": 0.5113,
+      "step": 6077
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5283091130467976,
+      "learning_rate": 7.0112611281561595e-06,
+      "loss": 0.494,
+      "step": 6078
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6262043520584162,
+      "learning_rate": 7.010316638376513e-06,
+      "loss": 0.5996,
+      "step": 6079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6629502911278257,
+      "learning_rate": 7.009372063024111e-06,
+      "loss": 0.5422,
+      "step": 6080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7604354893945472,
+      "learning_rate": 7.008427402139163e-06,
+      "loss": 0.6251,
+      "step": 6081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7047594686504335,
+      "learning_rate": 7.007482655761878e-06,
+      "loss": 0.5861,
+      "step": 6082
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8143829922209819,
+      "learning_rate": 7.006537823932471e-06,
+      "loss": 0.601,
+      "step": 6083
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5984028371930139,
+      "learning_rate": 7.005592906691161e-06,
+      "loss": 0.5225,
+      "step": 6084
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8038181984577378,
+      "learning_rate": 7.00464790407817e-06,
+      "loss": 0.612,
+      "step": 6085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7853545575807502,
+      "learning_rate": 7.003702816133724e-06,
+      "loss": 0.5401,
+      "step": 6086
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5447574378687547,
+      "learning_rate": 7.002757642898052e-06,
+      "loss": 0.5152,
+      "step": 6087
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7508248008587622,
+      "learning_rate": 7.001812384411386e-06,
+      "loss": 0.5991,
+      "step": 6088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6071495688858949,
+      "learning_rate": 7.000867040713964e-06,
+      "loss": 0.5257,
+      "step": 6089
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6580350397288754,
+      "learning_rate": 6.999921611846027e-06,
+      "loss": 0.5114,
+      "step": 6090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8296385344887232,
+      "learning_rate": 6.9989760978478176e-06,
+      "loss": 0.6806,
+      "step": 6091
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7362209605189728,
+      "learning_rate": 6.998030498759583e-06,
+      "loss": 0.5777,
+      "step": 6092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6859307273497655,
+      "learning_rate": 6.9970848146215744e-06,
+      "loss": 0.5421,
+      "step": 6093
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5844085917491771,
+      "learning_rate": 6.996139045474046e-06,
+      "loss": 0.532,
+      "step": 6094
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6592499049348985,
+      "learning_rate": 6.995193191357259e-06,
+      "loss": 0.563,
+      "step": 6095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5713688234567691,
+      "learning_rate": 6.994247252311473e-06,
+      "loss": 0.5138,
+      "step": 6096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6933565529333269,
+      "learning_rate": 6.993301228376955e-06,
+      "loss": 0.5335,
+      "step": 6097
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9250161177146866,
+      "learning_rate": 6.992355119593973e-06,
+      "loss": 0.622,
+      "step": 6098
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7224784839760994,
+      "learning_rate": 6.9914089260028e-06,
+      "loss": 0.5998,
+      "step": 6099
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.774091694077413,
+      "learning_rate": 6.990462647643713e-06,
+      "loss": 0.6345,
+      "step": 6100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5607801855546023,
+      "learning_rate": 6.989516284556992e-06,
+      "loss": 0.4911,
+      "step": 6101
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7074015476510126,
+      "learning_rate": 6.988569836782919e-06,
+      "loss": 0.4826,
+      "step": 6102
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8538333881835464,
+      "learning_rate": 6.9876233043617835e-06,
+      "loss": 0.6135,
+      "step": 6103
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6198661983849384,
+      "learning_rate": 6.986676687333874e-06,
+      "loss": 0.5423,
+      "step": 6104
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6908445499513478,
+      "learning_rate": 6.985729985739489e-06,
+      "loss": 0.526,
+      "step": 6105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6352547873271959,
+      "learning_rate": 6.984783199618922e-06,
+      "loss": 0.6285,
+      "step": 6106
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5914740537912223,
+      "learning_rate": 6.9838363290124776e-06,
+      "loss": 0.5254,
+      "step": 6107
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2801019088613952,
+      "learning_rate": 6.98288937396046e-06,
+      "loss": 0.6288,
+      "step": 6108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9335250808678409,
+      "learning_rate": 6.981942334503178e-06,
+      "loss": 0.6318,
+      "step": 6109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5849128006547969,
+      "learning_rate": 6.980995210680943e-06,
+      "loss": 0.5099,
+      "step": 6110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5681560596857154,
+      "learning_rate": 6.980048002534074e-06,
+      "loss": 0.5909,
+      "step": 6111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5871495885697342,
+      "learning_rate": 6.979100710102888e-06,
+      "loss": 0.485,
+      "step": 6112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8050561993364773,
+      "learning_rate": 6.97815333342771e-06,
+      "loss": 0.6033,
+      "step": 6113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7343342902711865,
+      "learning_rate": 6.9772058725488655e-06,
+      "loss": 0.6465,
+      "step": 6114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6408911840395797,
+      "learning_rate": 6.9762583275066855e-06,
+      "loss": 0.5082,
+      "step": 6115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7044054058239293,
+      "learning_rate": 6.975310698341503e-06,
+      "loss": 0.5339,
+      "step": 6116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5868739593970147,
+      "learning_rate": 6.974362985093656e-06,
+      "loss": 0.5566,
+      "step": 6117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6440622083548919,
+      "learning_rate": 6.973415187803486e-06,
+      "loss": 0.5513,
+      "step": 6118
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7066928954633134,
+      "learning_rate": 6.972467306511338e-06,
+      "loss": 0.5257,
+      "step": 6119
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7700288467201174,
+      "learning_rate": 6.97151934125756e-06,
+      "loss": 0.5991,
+      "step": 6120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5283508544117611,
+      "learning_rate": 6.970571292082504e-06,
+      "loss": 0.5175,
+      "step": 6121
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8470524703983034,
+      "learning_rate": 6.969623159026524e-06,
+      "loss": 0.6532,
+      "step": 6122
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6218391939761694,
+      "learning_rate": 6.96867494212998e-06,
+      "loss": 0.5331,
+      "step": 6123
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7473198316412153,
+      "learning_rate": 6.967726641433236e-06,
+      "loss": 0.6415,
+      "step": 6124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5365926497601979,
+      "learning_rate": 6.966778256976656e-06,
+      "loss": 0.4869,
+      "step": 6125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.914592815355564,
+      "learning_rate": 6.965829788800611e-06,
+      "loss": 0.6547,
+      "step": 6126
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6601173258279925,
+      "learning_rate": 6.964881236945474e-06,
+      "loss": 0.5607,
+      "step": 6127
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6843616194760515,
+      "learning_rate": 6.963932601451621e-06,
+      "loss": 0.5407,
+      "step": 6128
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7183738774638622,
+      "learning_rate": 6.962983882359433e-06,
+      "loss": 0.6158,
+      "step": 6129
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7680164649880659,
+      "learning_rate": 6.962035079709294e-06,
+      "loss": 0.6409,
+      "step": 6130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6684581297969855,
+      "learning_rate": 6.9610861935415915e-06,
+      "loss": 0.5631,
+      "step": 6131
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8363427677623987,
+      "learning_rate": 6.960137223896717e-06,
+      "loss": 0.6002,
+      "step": 6132
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7994319823281306,
+      "learning_rate": 6.9591881708150645e-06,
+      "loss": 0.5981,
+      "step": 6133
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6396602571762071,
+      "learning_rate": 6.958239034337032e-06,
+      "loss": 0.48,
+      "step": 6134
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9312708990308008,
+      "learning_rate": 6.95728981450302e-06,
+      "loss": 0.5926,
+      "step": 6135
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7540675079988468,
+      "learning_rate": 6.956340511353436e-06,
+      "loss": 0.6073,
+      "step": 6136
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7620861994525625,
+      "learning_rate": 6.95539112492869e-06,
+      "loss": 0.5923,
+      "step": 6137
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7773320261409922,
+      "learning_rate": 6.95444165526919e-06,
+      "loss": 0.6306,
+      "step": 6138
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6345931819255022,
+      "learning_rate": 6.953492102415355e-06,
+      "loss": 0.497,
+      "step": 6139
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7509745325030854,
+      "learning_rate": 6.9525424664076046e-06,
+      "loss": 0.6594,
+      "step": 6140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8573200746023567,
+      "learning_rate": 6.9515927472863595e-06,
+      "loss": 0.6061,
+      "step": 6141
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6054241820101158,
+      "learning_rate": 6.950642945092047e-06,
+      "loss": 0.5233,
+      "step": 6142
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8119505648001418,
+      "learning_rate": 6.9496930598651e-06,
+      "loss": 0.609,
+      "step": 6143
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6494219303020818,
+      "learning_rate": 6.948743091645949e-06,
+      "loss": 0.5748,
+      "step": 6144
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8789723812372376,
+      "learning_rate": 6.9477930404750304e-06,
+      "loss": 0.6145,
+      "step": 6145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6700853983739093,
+      "learning_rate": 6.9468429063927875e-06,
+      "loss": 0.5913,
+      "step": 6146
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7035478550174823,
+      "learning_rate": 6.945892689439664e-06,
+      "loss": 0.5895,
+      "step": 6147
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6381283890988907,
+      "learning_rate": 6.9449423896561055e-06,
+      "loss": 0.5454,
+      "step": 6148
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6237204562565154,
+      "learning_rate": 6.943992007082565e-06,
+      "loss": 0.5893,
+      "step": 6149
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8342605489853235,
+      "learning_rate": 6.943041541759496e-06,
+      "loss": 0.6199,
+      "step": 6150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6102616469181731,
+      "learning_rate": 6.94209099372736e-06,
+      "loss": 0.5517,
+      "step": 6151
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9253336997123479,
+      "learning_rate": 6.941140363026615e-06,
+      "loss": 0.5974,
+      "step": 6152
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.525117978333921,
+      "learning_rate": 6.940189649697728e-06,
+      "loss": 0.5245,
+      "step": 6153
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6832617034871739,
+      "learning_rate": 6.9392388537811675e-06,
+      "loss": 0.5946,
+      "step": 6154
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5714910527073446,
+      "learning_rate": 6.938287975317406e-06,
+      "loss": 0.5195,
+      "step": 6155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6361896105627376,
+      "learning_rate": 6.937337014346918e-06,
+      "loss": 0.6188,
+      "step": 6156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7734582689465479,
+      "learning_rate": 6.936385970910185e-06,
+      "loss": 0.5521,
+      "step": 6157
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7891876975678117,
+      "learning_rate": 6.9354348450476894e-06,
+      "loss": 0.6444,
+      "step": 6158
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7203664824765719,
+      "learning_rate": 6.934483636799918e-06,
+      "loss": 0.6059,
+      "step": 6159
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7088666589050271,
+      "learning_rate": 6.933532346207359e-06,
+      "loss": 0.565,
+      "step": 6160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0739967395205492,
+      "learning_rate": 6.932580973310507e-06,
+      "loss": 0.6433,
+      "step": 6161
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6721416735590974,
+      "learning_rate": 6.931629518149858e-06,
+      "loss": 0.5417,
+      "step": 6162
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8208870301459986,
+      "learning_rate": 6.930677980765913e-06,
+      "loss": 0.6081,
+      "step": 6163
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7918140216290176,
+      "learning_rate": 6.929726361199176e-06,
+      "loss": 0.6054,
+      "step": 6164
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6279839957932126,
+      "learning_rate": 6.928774659490155e-06,
+      "loss": 0.4815,
+      "step": 6165
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5384804074980266,
+      "learning_rate": 6.9278228756793594e-06,
+      "loss": 0.4604,
+      "step": 6166
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5964097962248348,
+      "learning_rate": 6.9268710098073044e-06,
+      "loss": 0.5459,
+      "step": 6167
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5860460691839549,
+      "learning_rate": 6.925919061914509e-06,
+      "loss": 0.4896,
+      "step": 6168
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6369287628579393,
+      "learning_rate": 6.924967032041493e-06,
+      "loss": 0.5444,
+      "step": 6169
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5809937042979003,
+      "learning_rate": 6.924014920228781e-06,
+      "loss": 0.5216,
+      "step": 6170
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.799396344707945,
+      "learning_rate": 6.923062726516902e-06,
+      "loss": 0.6009,
+      "step": 6171
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6579431336556247,
+      "learning_rate": 6.922110450946389e-06,
+      "loss": 0.5732,
+      "step": 6172
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9682610077255045,
+      "learning_rate": 6.921158093557776e-06,
+      "loss": 0.6369,
+      "step": 6173
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7393093470537417,
+      "learning_rate": 6.920205654391602e-06,
+      "loss": 0.6148,
+      "step": 6174
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7832586780668236,
+      "learning_rate": 6.9192531334884095e-06,
+      "loss": 0.6525,
+      "step": 6175
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7039747573022053,
+      "learning_rate": 6.918300530888744e-06,
+      "loss": 0.5907,
+      "step": 6176
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5574547025823923,
+      "learning_rate": 6.917347846633155e-06,
+      "loss": 0.5281,
+      "step": 6177
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.684420410474668,
+      "learning_rate": 6.916395080762196e-06,
+      "loss": 0.6124,
+      "step": 6178
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6630924917107729,
+      "learning_rate": 6.91544223331642e-06,
+      "loss": 0.5838,
+      "step": 6179
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0094688674188406,
+      "learning_rate": 6.914489304336391e-06,
+      "loss": 0.661,
+      "step": 6180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.693860965646948,
+      "learning_rate": 6.91353629386267e-06,
+      "loss": 0.5885,
+      "step": 6181
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6045233868854597,
+      "learning_rate": 6.912583201935823e-06,
+      "loss": 0.4921,
+      "step": 6182
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5976208342295841,
+      "learning_rate": 6.911630028596421e-06,
+      "loss": 0.5161,
+      "step": 6183
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6133173503245882,
+      "learning_rate": 6.9106767738850386e-06,
+      "loss": 0.4978,
+      "step": 6184
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.710504803794944,
+      "learning_rate": 6.909723437842249e-06,
+      "loss": 0.628,
+      "step": 6185
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7743011024412121,
+      "learning_rate": 6.908770020508637e-06,
+      "loss": 0.6303,
+      "step": 6186
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6856054433042441,
+      "learning_rate": 6.907816521924785e-06,
+      "loss": 0.557,
+      "step": 6187
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.781354247419235,
+      "learning_rate": 6.90686294213128e-06,
+      "loss": 0.5982,
+      "step": 6188
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.754877037475734,
+      "learning_rate": 6.905909281168713e-06,
+      "loss": 0.6367,
+      "step": 6189
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7514741435184724,
+      "learning_rate": 6.904955539077678e-06,
+      "loss": 0.5725,
+      "step": 6190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8216087262393915,
+      "learning_rate": 6.904001715898773e-06,
+      "loss": 0.6266,
+      "step": 6191
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7563601759227156,
+      "learning_rate": 6.903047811672599e-06,
+      "loss": 0.5646,
+      "step": 6192
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6395881979251259,
+      "learning_rate": 6.9020938264397615e-06,
+      "loss": 0.4992,
+      "step": 6193
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5733684592744743,
+      "learning_rate": 6.901139760240868e-06,
+      "loss": 0.4656,
+      "step": 6194
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5579104745176718,
+      "learning_rate": 6.90018561311653e-06,
+      "loss": 0.529,
+      "step": 6195
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8436171976687403,
+      "learning_rate": 6.899231385107364e-06,
+      "loss": 0.619,
+      "step": 6196
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5820596508606891,
+      "learning_rate": 6.8982770762539845e-06,
+      "loss": 0.5333,
+      "step": 6197
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8198069024579641,
+      "learning_rate": 6.897322686597016e-06,
+      "loss": 0.5821,
+      "step": 6198
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5074964758574475,
+      "learning_rate": 6.8963682161770854e-06,
+      "loss": 0.4855,
+      "step": 6199
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5053268102071425,
+      "learning_rate": 6.895413665034819e-06,
+      "loss": 0.461,
+      "step": 6200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7541032712826765,
+      "learning_rate": 6.89445903321085e-06,
+      "loss": 0.577,
+      "step": 6201
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7463920325018282,
+      "learning_rate": 6.893504320745814e-06,
+      "loss": 0.6272,
+      "step": 6202
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6227189089476501,
+      "learning_rate": 6.892549527680348e-06,
+      "loss": 0.5534,
+      "step": 6203
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6424807333780551,
+      "learning_rate": 6.891594654055098e-06,
+      "loss": 0.5329,
+      "step": 6204
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7734804845782416,
+      "learning_rate": 6.8906396999107085e-06,
+      "loss": 0.5427,
+      "step": 6205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.742196668913134,
+      "learning_rate": 6.889684665287828e-06,
+      "loss": 0.6229,
+      "step": 6206
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8613033288837254,
+      "learning_rate": 6.88872955022711e-06,
+      "loss": 0.679,
+      "step": 6207
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6666564855842763,
+      "learning_rate": 6.88777435476921e-06,
+      "loss": 0.5233,
+      "step": 6208
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8033239955118487,
+      "learning_rate": 6.88681907895479e-06,
+      "loss": 0.6039,
+      "step": 6209
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7021445627143932,
+      "learning_rate": 6.8858637228245086e-06,
+      "loss": 0.6238,
+      "step": 6210
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6665255342789965,
+      "learning_rate": 6.884908286419036e-06,
+      "loss": 0.5371,
+      "step": 6211
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5943111470724661,
+      "learning_rate": 6.8839527697790405e-06,
+      "loss": 0.5069,
+      "step": 6212
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5626700224449825,
+      "learning_rate": 6.882997172945198e-06,
+      "loss": 0.509,
+      "step": 6213
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7003198306754069,
+      "learning_rate": 6.882041495958181e-06,
+      "loss": 0.4998,
+      "step": 6214
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5864565904943859,
+      "learning_rate": 6.881085738858673e-06,
+      "loss": 0.513,
+      "step": 6215
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5634487255844925,
+      "learning_rate": 6.880129901687356e-06,
+      "loss": 0.4916,
+      "step": 6216
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5806470534231932,
+      "learning_rate": 6.879173984484915e-06,
+      "loss": 0.5266,
+      "step": 6217
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7388452903608196,
+      "learning_rate": 6.878217987292044e-06,
+      "loss": 0.626,
+      "step": 6218
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7120962507420958,
+      "learning_rate": 6.8772619101494356e-06,
+      "loss": 0.545,
+      "step": 6219
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5667017214524706,
+      "learning_rate": 6.876305753097786e-06,
+      "loss": 0.4413,
+      "step": 6220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9126339519742783,
+      "learning_rate": 6.875349516177796e-06,
+      "loss": 0.607,
+      "step": 6221
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6508330591251228,
+      "learning_rate": 6.87439319943017e-06,
+      "loss": 0.5155,
+      "step": 6222
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.559887847798657,
+      "learning_rate": 6.873436802895617e-06,
+      "loss": 0.5099,
+      "step": 6223
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7936099331457016,
+      "learning_rate": 6.872480326614844e-06,
+      "loss": 0.5903,
+      "step": 6224
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.610167109177135,
+      "learning_rate": 6.871523770628568e-06,
+      "loss": 0.5426,
+      "step": 6225
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8166665705034846,
+      "learning_rate": 6.870567134977505e-06,
+      "loss": 0.6252,
+      "step": 6226
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7287601280290925,
+      "learning_rate": 6.8696104197023776e-06,
+      "loss": 0.6008,
+      "step": 6227
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.601670641793849,
+      "learning_rate": 6.868653624843906e-06,
+      "loss": 0.5068,
+      "step": 6228
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7617357785708877,
+      "learning_rate": 6.8676967504428235e-06,
+      "loss": 0.61,
+      "step": 6229
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7307420799092397,
+      "learning_rate": 6.866739796539859e-06,
+      "loss": 0.6041,
+      "step": 6230
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5876454093073531,
+      "learning_rate": 6.865782763175746e-06,
+      "loss": 0.4629,
+      "step": 6231
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7987521085472857,
+      "learning_rate": 6.864825650391223e-06,
+      "loss": 0.6566,
+      "step": 6232
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7055230017174057,
+      "learning_rate": 6.863868458227031e-06,
+      "loss": 0.5847,
+      "step": 6233
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7088863941530662,
+      "learning_rate": 6.862911186723914e-06,
+      "loss": 0.5952,
+      "step": 6234
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7315691821038331,
+      "learning_rate": 6.861953835922621e-06,
+      "loss": 0.6058,
+      "step": 6235
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7202830970738314,
+      "learning_rate": 6.860996405863905e-06,
+      "loss": 0.5745,
+      "step": 6236
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7226304380598846,
+      "learning_rate": 6.860038896588517e-06,
+      "loss": 0.5845,
+      "step": 6237
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8105305245141821,
+      "learning_rate": 6.859081308137217e-06,
+      "loss": 0.6127,
+      "step": 6238
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8045750980169796,
+      "learning_rate": 6.858123640550768e-06,
+      "loss": 0.5539,
+      "step": 6239
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6142025774557623,
+      "learning_rate": 6.8571658938699325e-06,
+      "loss": 0.5098,
+      "step": 6240
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7187259568110902,
+      "learning_rate": 6.856208068135479e-06,
+      "loss": 0.6084,
+      "step": 6241
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7185962916088359,
+      "learning_rate": 6.855250163388181e-06,
+      "loss": 0.5968,
+      "step": 6242
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8012193986950965,
+      "learning_rate": 6.854292179668811e-06,
+      "loss": 0.5973,
+      "step": 6243
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6125550453861028,
+      "learning_rate": 6.853334117018149e-06,
+      "loss": 0.4825,
+      "step": 6244
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.749544309407826,
+      "learning_rate": 6.852375975476975e-06,
+      "loss": 0.5419,
+      "step": 6245
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5020314314941644,
+      "learning_rate": 6.851417755086076e-06,
+      "loss": 0.4914,
+      "step": 6246
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7697754095881776,
+      "learning_rate": 6.850459455886238e-06,
+      "loss": 0.6191,
+      "step": 6247
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5241607131772357,
+      "learning_rate": 6.8495010779182555e-06,
+      "loss": 0.5187,
+      "step": 6248
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5733990169862537,
+      "learning_rate": 6.848542621222922e-06,
+      "loss": 0.5339,
+      "step": 6249
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9785630042083076,
+      "learning_rate": 6.847584085841037e-06,
+      "loss": 0.6041,
+      "step": 6250
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6649692871869353,
+      "learning_rate": 6.846625471813402e-06,
+      "loss": 0.5882,
+      "step": 6251
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8112827043327446,
+      "learning_rate": 6.84566677918082e-06,
+      "loss": 0.6561,
+      "step": 6252
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7365231857471037,
+      "learning_rate": 6.8447080079841e-06,
+      "loss": 0.5633,
+      "step": 6253
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5696004897404208,
+      "learning_rate": 6.843749158264057e-06,
+      "loss": 0.5103,
+      "step": 6254
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6108554199444037,
+      "learning_rate": 6.842790230061504e-06,
+      "loss": 0.5826,
+      "step": 6255
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6850630542323604,
+      "learning_rate": 6.841831223417259e-06,
+      "loss": 0.5647,
+      "step": 6256
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7699240680149971,
+      "learning_rate": 6.840872138372145e-06,
+      "loss": 0.6505,
+      "step": 6257
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5776650376463175,
+      "learning_rate": 6.8399129749669866e-06,
+      "loss": 0.533,
+      "step": 6258
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5806014451399877,
+      "learning_rate": 6.838953733242611e-06,
+      "loss": 0.5068,
+      "step": 6259
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6530780661175465,
+      "learning_rate": 6.837994413239854e-06,
+      "loss": 0.5724,
+      "step": 6260
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8350028438034934,
+      "learning_rate": 6.837035014999546e-06,
+      "loss": 0.5839,
+      "step": 6261
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7615034528067307,
+      "learning_rate": 6.836075538562529e-06,
+      "loss": 0.602,
+      "step": 6262
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5822667151176976,
+      "learning_rate": 6.8351159839696436e-06,
+      "loss": 0.5173,
+      "step": 6263
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.755752935684026,
+      "learning_rate": 6.8341563512617335e-06,
+      "loss": 0.5996,
+      "step": 6264
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7101963630379892,
+      "learning_rate": 6.833196640479649e-06,
+      "loss": 0.5841,
+      "step": 6265
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6852151021143676,
+      "learning_rate": 6.832236851664242e-06,
+      "loss": 0.5554,
+      "step": 6266
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6015082261462827,
+      "learning_rate": 6.831276984856369e-06,
+      "loss": 0.497,
+      "step": 6267
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6176942236606779,
+      "learning_rate": 6.830317040096886e-06,
+      "loss": 0.4986,
+      "step": 6268
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7603440133190384,
+      "learning_rate": 6.8293570174266535e-06,
+      "loss": 0.6027,
+      "step": 6269
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.637943257089538,
+      "learning_rate": 6.828396916886539e-06,
+      "loss": 0.5489,
+      "step": 6270
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7500199822387734,
+      "learning_rate": 6.827436738517412e-06,
+      "loss": 0.5838,
+      "step": 6271
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6582165205043784,
+      "learning_rate": 6.8264764823601404e-06,
+      "loss": 0.5221,
+      "step": 6272
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7077572872091533,
+      "learning_rate": 6.8255161484556045e-06,
+      "loss": 0.586,
+      "step": 6273
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7533088625713023,
+      "learning_rate": 6.8245557368446776e-06,
+      "loss": 0.6029,
+      "step": 6274
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7037599100165665,
+      "learning_rate": 6.823595247568244e-06,
+      "loss": 0.5991,
+      "step": 6275
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6571375999275945,
+      "learning_rate": 6.8226346806671885e-06,
+      "loss": 0.529,
+      "step": 6276
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6469866110133251,
+      "learning_rate": 6.821674036182399e-06,
+      "loss": 0.5267,
+      "step": 6277
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5644371973477075,
+      "learning_rate": 6.820713314154765e-06,
+      "loss": 0.5203,
+      "step": 6278
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5959305779758147,
+      "learning_rate": 6.819752514625184e-06,
+      "loss": 0.5317,
+      "step": 6279
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7102077211387456,
+      "learning_rate": 6.8187916376345555e-06,
+      "loss": 0.6044,
+      "step": 6280
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5544538366449274,
+      "learning_rate": 6.817830683223778e-06,
+      "loss": 0.5082,
+      "step": 6281
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8184035636411395,
+      "learning_rate": 6.816869651433757e-06,
+      "loss": 0.6817,
+      "step": 6282
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7213404586909871,
+      "learning_rate": 6.815908542305402e-06,
+      "loss": 0.5216,
+      "step": 6283
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7223103989692812,
+      "learning_rate": 6.814947355879622e-06,
+      "loss": 0.5678,
+      "step": 6284
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7481573018533019,
+      "learning_rate": 6.8139860921973335e-06,
+      "loss": 0.5798,
+      "step": 6285
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6012225613001245,
+      "learning_rate": 6.813024751299453e-06,
+      "loss": 0.5014,
+      "step": 6286
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6577919408740096,
+      "learning_rate": 6.812063333226903e-06,
+      "loss": 0.5676,
+      "step": 6287
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5780130804933448,
+      "learning_rate": 6.8111018380206075e-06,
+      "loss": 0.5257,
+      "step": 6288
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6037215296113814,
+      "learning_rate": 6.8101402657214934e-06,
+      "loss": 0.5057,
+      "step": 6289
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6517998028216473,
+      "learning_rate": 6.809178616370492e-06,
+      "loss": 0.5293,
+      "step": 6290
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6471955927017133,
+      "learning_rate": 6.8082168900085385e-06,
+      "loss": 0.525,
+      "step": 6291
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8668090658652021,
+      "learning_rate": 6.807255086676572e-06,
+      "loss": 0.6323,
+      "step": 6292
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7283029204199846,
+      "learning_rate": 6.80629320641553e-06,
+      "loss": 0.6375,
+      "step": 6293
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.777387050154529,
+      "learning_rate": 6.805331249266359e-06,
+      "loss": 0.6458,
+      "step": 6294
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6352214781589633,
+      "learning_rate": 6.8043692152700056e-06,
+      "loss": 0.5185,
+      "step": 6295
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.684343183397395,
+      "learning_rate": 6.803407104467421e-06,
+      "loss": 0.6893,
+      "step": 6296
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7276600518156257,
+      "learning_rate": 6.802444916899558e-06,
+      "loss": 0.5878,
+      "step": 6297
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6796942134435697,
+      "learning_rate": 6.801482652607375e-06,
+      "loss": 0.5382,
+      "step": 6298
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6201585270623763,
+      "learning_rate": 6.800520311631833e-06,
+      "loss": 0.5295,
+      "step": 6299
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7462001506166249,
+      "learning_rate": 6.799557894013894e-06,
+      "loss": 0.6015,
+      "step": 6300
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6913324373138605,
+      "learning_rate": 6.798595399794527e-06,
+      "loss": 0.6534,
+      "step": 6301
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8357334549675459,
+      "learning_rate": 6.7976328290147e-06,
+      "loss": 0.6185,
+      "step": 6302
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.546093348874749,
+      "learning_rate": 6.796670181715388e-06,
+      "loss": 0.474,
+      "step": 6303
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7175699288559444,
+      "learning_rate": 6.795707457937568e-06,
+      "loss": 0.585,
+      "step": 6304
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6450632005756834,
+      "learning_rate": 6.794744657722221e-06,
+      "loss": 0.5129,
+      "step": 6305
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7124027743503827,
+      "learning_rate": 6.793781781110328e-06,
+      "loss": 0.6108,
+      "step": 6306
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8872452299574621,
+      "learning_rate": 6.792818828142876e-06,
+      "loss": 0.6345,
+      "step": 6307
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7730338132466026,
+      "learning_rate": 6.791855798860857e-06,
+      "loss": 0.5703,
+      "step": 6308
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6589697691999539,
+      "learning_rate": 6.790892693305261e-06,
+      "loss": 0.5561,
+      "step": 6309
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6191895272824437,
+      "learning_rate": 6.789929511517087e-06,
+      "loss": 0.5128,
+      "step": 6310
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7340550573357525,
+      "learning_rate": 6.788966253537333e-06,
+      "loss": 0.6193,
+      "step": 6311
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6602665420114553,
+      "learning_rate": 6.788002919407003e-06,
+      "loss": 0.4922,
+      "step": 6312
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6551173853319442,
+      "learning_rate": 6.7870395091671014e-06,
+      "loss": 0.5302,
+      "step": 6313
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8288366529652099,
+      "learning_rate": 6.78607602285864e-06,
+      "loss": 0.6447,
+      "step": 6314
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8531539098497349,
+      "learning_rate": 6.785112460522627e-06,
+      "loss": 0.6653,
+      "step": 6315
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7759389167907225,
+      "learning_rate": 6.784148822200084e-06,
+      "loss": 0.6719,
+      "step": 6316
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6430947513586196,
+      "learning_rate": 6.783185107932025e-06,
+      "loss": 0.4969,
+      "step": 6317
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6864824543720452,
+      "learning_rate": 6.782221317759475e-06,
+      "loss": 0.5667,
+      "step": 6318
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5712421829392823,
+      "learning_rate": 6.78125745172346e-06,
+      "loss": 0.5097,
+      "step": 6319
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9133388746218416,
+      "learning_rate": 6.7802935098650055e-06,
+      "loss": 0.6281,
+      "step": 6320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8147174738533315,
+      "learning_rate": 6.779329492225144e-06,
+      "loss": 0.6049,
+      "step": 6321
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6203870899888008,
+      "learning_rate": 6.778365398844915e-06,
+      "loss": 0.5212,
+      "step": 6322
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6220443670644604,
+      "learning_rate": 6.7774012297653525e-06,
+      "loss": 0.5344,
+      "step": 6323
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7281004215424857,
+      "learning_rate": 6.776436985027499e-06,
+      "loss": 0.5908,
+      "step": 6324
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.766516496697392,
+      "learning_rate": 6.7754726646724e-06,
+      "loss": 0.5989,
+      "step": 6325
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8734819449120286,
+      "learning_rate": 6.774508268741105e-06,
+      "loss": 0.5812,
+      "step": 6326
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6165868605416652,
+      "learning_rate": 6.77354379727466e-06,
+      "loss": 0.556,
+      "step": 6327
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7544123996488468,
+      "learning_rate": 6.772579250314126e-06,
+      "loss": 0.5535,
+      "step": 6328
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6843396353768542,
+      "learning_rate": 6.771614627900556e-06,
+      "loss": 0.5076,
+      "step": 6329
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7335055256580995,
+      "learning_rate": 6.770649930075013e-06,
+      "loss": 0.626,
+      "step": 6330
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8637164624456584,
+      "learning_rate": 6.76968515687856e-06,
+      "loss": 0.6601,
+      "step": 6331
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6383292611934841,
+      "learning_rate": 6.768720308352267e-06,
+      "loss": 0.5465,
+      "step": 6332
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0248782307901643,
+      "learning_rate": 6.767755384537202e-06,
+      "loss": 0.621,
+      "step": 6333
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.779170201696406,
+      "learning_rate": 6.766790385474436e-06,
+      "loss": 0.6268,
+      "step": 6334
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6401784606901068,
+      "learning_rate": 6.765825311205052e-06,
+      "loss": 0.5438,
+      "step": 6335
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7786211418180801,
+      "learning_rate": 6.764860161770128e-06,
+      "loss": 0.6498,
+      "step": 6336
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6726624664743807,
+      "learning_rate": 6.7638949372107455e-06,
+      "loss": 0.5716,
+      "step": 6337
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6774421434867293,
+      "learning_rate": 6.762929637567992e-06,
+      "loss": 0.5998,
+      "step": 6338
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5738097360089144,
+      "learning_rate": 6.761964262882957e-06,
+      "loss": 0.5201,
+      "step": 6339
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.578057836464607,
+      "learning_rate": 6.760998813196735e-06,
+      "loss": 0.5358,
+      "step": 6340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9411607403286338,
+      "learning_rate": 6.760033288550419e-06,
+      "loss": 0.6528,
+      "step": 6341
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7901292814985689,
+      "learning_rate": 6.759067688985111e-06,
+      "loss": 0.6205,
+      "step": 6342
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7859494662537901,
+      "learning_rate": 6.758102014541914e-06,
+      "loss": 0.6246,
+      "step": 6343
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5360097350128014,
+      "learning_rate": 6.757136265261931e-06,
+      "loss": 0.4813,
+      "step": 6344
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6384464118020083,
+      "learning_rate": 6.756170441186273e-06,
+      "loss": 0.575,
+      "step": 6345
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7887056311658751,
+      "learning_rate": 6.755204542356051e-06,
+      "loss": 0.6061,
+      "step": 6346
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7973621020106734,
+      "learning_rate": 6.754238568812379e-06,
+      "loss": 0.6175,
+      "step": 6347
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5730633541510033,
+      "learning_rate": 6.753272520596379e-06,
+      "loss": 0.5138,
+      "step": 6348
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.647216245800709,
+      "learning_rate": 6.752306397749168e-06,
+      "loss": 0.5691,
+      "step": 6349
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5954325889514445,
+      "learning_rate": 6.751340200311875e-06,
+      "loss": 0.5111,
+      "step": 6350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6657911872799888,
+      "learning_rate": 6.7503739283256275e-06,
+      "loss": 0.5294,
+      "step": 6351
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5328784579864676,
+      "learning_rate": 6.749407581831553e-06,
+      "loss": 0.5258,
+      "step": 6352
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8368523370346123,
+      "learning_rate": 6.748441160870788e-06,
+      "loss": 0.6113,
+      "step": 6353
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5718359238750406,
+      "learning_rate": 6.74747466548447e-06,
+      "loss": 0.5504,
+      "step": 6354
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7893892783212353,
+      "learning_rate": 6.746508095713742e-06,
+      "loss": 0.6742,
+      "step": 6355
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6334354873352155,
+      "learning_rate": 6.745541451599743e-06,
+      "loss": 0.576,
+      "step": 6356
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6671874039448502,
+      "learning_rate": 6.7445747331836235e-06,
+      "loss": 0.5022,
+      "step": 6357
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5682849379380871,
+      "learning_rate": 6.743607940506531e-06,
+      "loss": 0.5301,
+      "step": 6358
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6529245815969871,
+      "learning_rate": 6.742641073609621e-06,
+      "loss": 0.5456,
+      "step": 6359
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5529149987318653,
+      "learning_rate": 6.74167413253405e-06,
+      "loss": 0.5154,
+      "step": 6360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6271718459165078,
+      "learning_rate": 6.740707117320976e-06,
+      "loss": 0.511,
+      "step": 6361
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8792267567828937,
+      "learning_rate": 6.739740028011564e-06,
+      "loss": 0.6005,
+      "step": 6362
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6326283319976258,
+      "learning_rate": 6.738772864646976e-06,
+      "loss": 0.5839,
+      "step": 6363
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7789312932868144,
+      "learning_rate": 6.737805627268385e-06,
+      "loss": 0.5517,
+      "step": 6364
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8044445950198794,
+      "learning_rate": 6.73683831591696e-06,
+      "loss": 0.6369,
+      "step": 6365
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5872765022955363,
+      "learning_rate": 6.73587093063388e-06,
+      "loss": 0.5741,
+      "step": 6366
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6077595396691541,
+      "learning_rate": 6.734903471460321e-06,
+      "loss": 0.5523,
+      "step": 6367
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6037433871943244,
+      "learning_rate": 6.733935938437466e-06,
+      "loss": 0.5268,
+      "step": 6368
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8342328555210821,
+      "learning_rate": 6.732968331606498e-06,
+      "loss": 0.6046,
+      "step": 6369
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.623905236072365,
+      "learning_rate": 6.732000651008606e-06,
+      "loss": 0.5542,
+      "step": 6370
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5152553373721863,
+      "learning_rate": 6.731032896684979e-06,
+      "loss": 0.4848,
+      "step": 6371
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6491968961014971,
+      "learning_rate": 6.730065068676816e-06,
+      "loss": 0.5371,
+      "step": 6372
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6969049736847973,
+      "learning_rate": 6.72909716702531e-06,
+      "loss": 0.5595,
+      "step": 6373
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5510741188143825,
+      "learning_rate": 6.728129191771664e-06,
+      "loss": 0.4857,
+      "step": 6374
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7451068437304299,
+      "learning_rate": 6.727161142957081e-06,
+      "loss": 0.6152,
+      "step": 6375
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6298200907591202,
+      "learning_rate": 6.726193020622766e-06,
+      "loss": 0.49,
+      "step": 6376
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6542965102588401,
+      "learning_rate": 6.72522482480993e-06,
+      "loss": 0.5584,
+      "step": 6377
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7509972788597397,
+      "learning_rate": 6.724256555559787e-06,
+      "loss": 0.6335,
+      "step": 6378
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6254214657372537,
+      "learning_rate": 6.723288212913553e-06,
+      "loss": 0.5257,
+      "step": 6379
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7173655135127083,
+      "learning_rate": 6.722319796912446e-06,
+      "loss": 0.5995,
+      "step": 6380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8969418779895143,
+      "learning_rate": 6.721351307597689e-06,
+      "loss": 0.6029,
+      "step": 6381
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7638516192884462,
+      "learning_rate": 6.720382745010507e-06,
+      "loss": 0.5027,
+      "step": 6382
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6582976704035635,
+      "learning_rate": 6.71941410919213e-06,
+      "loss": 0.4851,
+      "step": 6383
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1897043957818842,
+      "learning_rate": 6.718445400183789e-06,
+      "loss": 0.6378,
+      "step": 6384
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7208578914676788,
+      "learning_rate": 6.717476618026717e-06,
+      "loss": 0.5765,
+      "step": 6385
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5721751519254793,
+      "learning_rate": 6.716507762762155e-06,
+      "loss": 0.4664,
+      "step": 6386
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7802442509445575,
+      "learning_rate": 6.7155388344313434e-06,
+      "loss": 0.6487,
+      "step": 6387
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.882001307796168,
+      "learning_rate": 6.714569833075524e-06,
+      "loss": 0.6357,
+      "step": 6388
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.879967952624162,
+      "learning_rate": 6.713600758735946e-06,
+      "loss": 0.5818,
+      "step": 6389
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5626014171327017,
+      "learning_rate": 6.71263161145386e-06,
+      "loss": 0.5003,
+      "step": 6390
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7976670591117455,
+      "learning_rate": 6.7116623912705196e-06,
+      "loss": 0.6132,
+      "step": 6391
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5724625636422039,
+      "learning_rate": 6.710693098227181e-06,
+      "loss": 0.4979,
+      "step": 6392
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7429685766744689,
+      "learning_rate": 6.709723732365103e-06,
+      "loss": 0.5725,
+      "step": 6393
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5828556645578636,
+      "learning_rate": 6.70875429372555e-06,
+      "loss": 0.5157,
+      "step": 6394
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6320381571843299,
+      "learning_rate": 6.707784782349787e-06,
+      "loss": 0.5201,
+      "step": 6395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6159010265565266,
+      "learning_rate": 6.706815198279082e-06,
+      "loss": 0.5008,
+      "step": 6396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5737740421180026,
+      "learning_rate": 6.7058455415547085e-06,
+      "loss": 0.4834,
+      "step": 6397
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5789861003246338,
+      "learning_rate": 6.704875812217942e-06,
+      "loss": 0.472,
+      "step": 6398
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6760633839978232,
+      "learning_rate": 6.7039060103100605e-06,
+      "loss": 0.6028,
+      "step": 6399
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7588558279356667,
+      "learning_rate": 6.702936135872344e-06,
+      "loss": 0.6653,
+      "step": 6400
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5963435157206481,
+      "learning_rate": 6.70196618894608e-06,
+      "loss": 0.5475,
+      "step": 6401
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.736899258733208,
+      "learning_rate": 6.700996169572553e-06,
+      "loss": 0.5398,
+      "step": 6402
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8047954247923363,
+      "learning_rate": 6.700026077793052e-06,
+      "loss": 0.6701,
+      "step": 6403
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7811606989006268,
+      "learning_rate": 6.699055913648877e-06,
+      "loss": 0.5097,
+      "step": 6404
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8208408967904837,
+      "learning_rate": 6.69808567718132e-06,
+      "loss": 0.6005,
+      "step": 6405
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6951735888945008,
+      "learning_rate": 6.6971153684316815e-06,
+      "loss": 0.5977,
+      "step": 6406
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7181001260605885,
+      "learning_rate": 6.696144987441265e-06,
+      "loss": 0.5459,
+      "step": 6407
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2468873166070857,
+      "learning_rate": 6.695174534251377e-06,
+      "loss": 0.6572,
+      "step": 6408
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7854443515098432,
+      "learning_rate": 6.694204008903326e-06,
+      "loss": 0.6118,
+      "step": 6409
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5717858423249883,
+      "learning_rate": 6.693233411438424e-06,
+      "loss": 0.4987,
+      "step": 6410
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7099312229948749,
+      "learning_rate": 6.692262741897988e-06,
+      "loss": 0.594,
+      "step": 6411
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8691331635780105,
+      "learning_rate": 6.691292000323333e-06,
+      "loss": 0.6265,
+      "step": 6412
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6657924821602142,
+      "learning_rate": 6.690321186755783e-06,
+      "loss": 0.4977,
+      "step": 6413
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8928952064216477,
+      "learning_rate": 6.68935030123666e-06,
+      "loss": 0.5673,
+      "step": 6414
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6431463191041895,
+      "learning_rate": 6.688379343807294e-06,
+      "loss": 0.5669,
+      "step": 6415
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.891696782672777,
+      "learning_rate": 6.687408314509015e-06,
+      "loss": 0.6501,
+      "step": 6416
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7976500124486192,
+      "learning_rate": 6.6864372133831565e-06,
+      "loss": 0.5879,
+      "step": 6417
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8207467405287292,
+      "learning_rate": 6.685466040471055e-06,
+      "loss": 0.6457,
+      "step": 6418
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8132338002114149,
+      "learning_rate": 6.6844947958140495e-06,
+      "loss": 0.5782,
+      "step": 6419
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5865131703071755,
+      "learning_rate": 6.683523479453482e-06,
+      "loss": 0.5329,
+      "step": 6420
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7642471230476104,
+      "learning_rate": 6.682552091430702e-06,
+      "loss": 0.6505,
+      "step": 6421
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6269724046919594,
+      "learning_rate": 6.681580631787055e-06,
+      "loss": 0.5537,
+      "step": 6422
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.787866405621647,
+      "learning_rate": 6.680609100563895e-06,
+      "loss": 0.6021,
+      "step": 6423
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.63042795399231,
+      "learning_rate": 6.679637497802576e-06,
+      "loss": 0.573,
+      "step": 6424
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5839167676917506,
+      "learning_rate": 6.6786658235444545e-06,
+      "loss": 0.4792,
+      "step": 6425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8759054067242198,
+      "learning_rate": 6.677694077830895e-06,
+      "loss": 0.6558,
+      "step": 6426
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6307238457096034,
+      "learning_rate": 6.676722260703257e-06,
+      "loss": 0.5357,
+      "step": 6427
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9370918199493273,
+      "learning_rate": 6.675750372202914e-06,
+      "loss": 0.6828,
+      "step": 6428
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8054200309083763,
+      "learning_rate": 6.674778412371231e-06,
+      "loss": 0.5958,
+      "step": 6429
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7846811580356037,
+      "learning_rate": 6.673806381249582e-06,
+      "loss": 0.6207,
+      "step": 6430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5581675605256007,
+      "learning_rate": 6.6728342788793455e-06,
+      "loss": 0.5634,
+      "step": 6431
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7101407156229939,
+      "learning_rate": 6.671862105301898e-06,
+      "loss": 0.5452,
+      "step": 6432
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7614938372698999,
+      "learning_rate": 6.670889860558623e-06,
+      "loss": 0.4759,
+      "step": 6433
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8168065779181795,
+      "learning_rate": 6.669917544690908e-06,
+      "loss": 0.5904,
+      "step": 6434
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6750045326616718,
+      "learning_rate": 6.668945157740139e-06,
+      "loss": 0.4993,
+      "step": 6435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7930097513448877,
+      "learning_rate": 6.667972699747707e-06,
+      "loss": 0.6017,
+      "step": 6436
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5277232096528562,
+      "learning_rate": 6.667000170755007e-06,
+      "loss": 0.4836,
+      "step": 6437
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5934540690664346,
+      "learning_rate": 6.666027570803437e-06,
+      "loss": 0.544,
+      "step": 6438
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5690663884221651,
+      "learning_rate": 6.665054899934397e-06,
+      "loss": 0.5077,
+      "step": 6439
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6735525718196225,
+      "learning_rate": 6.664082158189291e-06,
+      "loss": 0.5782,
+      "step": 6440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7205738057595237,
+      "learning_rate": 6.663109345609525e-06,
+      "loss": 0.5601,
+      "step": 6441
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7535096231438502,
+      "learning_rate": 6.662136462236509e-06,
+      "loss": 0.6243,
+      "step": 6442
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6220989097344559,
+      "learning_rate": 6.661163508111655e-06,
+      "loss": 0.5077,
+      "step": 6443
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7335877244020317,
+      "learning_rate": 6.66019048327638e-06,
+      "loss": 0.594,
+      "step": 6444
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5847035984154572,
+      "learning_rate": 6.659217387772099e-06,
+      "loss": 0.5017,
+      "step": 6445
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6953747917685108,
+      "learning_rate": 6.658244221640238e-06,
+      "loss": 0.5715,
+      "step": 6446
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7203864369924682,
+      "learning_rate": 6.657270984922217e-06,
+      "loss": 0.5861,
+      "step": 6447
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5187896783444225,
+      "learning_rate": 6.6562976776594686e-06,
+      "loss": 0.5062,
+      "step": 6448
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7099647596179366,
+      "learning_rate": 6.65532429989342e-06,
+      "loss": 0.5598,
+      "step": 6449
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7002824694552254,
+      "learning_rate": 6.654350851665505e-06,
+      "loss": 0.5226,
+      "step": 6450
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.721513877311857,
+      "learning_rate": 6.65337733301716e-06,
+      "loss": 0.546,
+      "step": 6451
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5634849603066018,
+      "learning_rate": 6.652403743989827e-06,
+      "loss": 0.5039,
+      "step": 6452
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.569399846285577,
+      "learning_rate": 6.651430084624947e-06,
+      "loss": 0.4602,
+      "step": 6453
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6769632673894282,
+      "learning_rate": 6.650456354963966e-06,
+      "loss": 0.5568,
+      "step": 6454
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6914939967947034,
+      "learning_rate": 6.64948255504833e-06,
+      "loss": 0.5609,
+      "step": 6455
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8084550059355045,
+      "learning_rate": 6.648508684919495e-06,
+      "loss": 0.6309,
+      "step": 6456
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5736143024332176,
+      "learning_rate": 6.647534744618911e-06,
+      "loss": 0.4859,
+      "step": 6457
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5459911748431158,
+      "learning_rate": 6.646560734188039e-06,
+      "loss": 0.5395,
+      "step": 6458
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.647497937146447,
+      "learning_rate": 6.645586653668337e-06,
+      "loss": 0.5455,
+      "step": 6459
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7416950392981845,
+      "learning_rate": 6.644612503101271e-06,
+      "loss": 0.5421,
+      "step": 6460
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.769850560611796,
+      "learning_rate": 6.643638282528306e-06,
+      "loss": 0.6333,
+      "step": 6461
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.787546334563346,
+      "learning_rate": 6.642663991990911e-06,
+      "loss": 0.6094,
+      "step": 6462
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6316002600992692,
+      "learning_rate": 6.641689631530559e-06,
+      "loss": 0.5062,
+      "step": 6463
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7174694295250796,
+      "learning_rate": 6.640715201188727e-06,
+      "loss": 0.6388,
+      "step": 6464
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7925445165429377,
+      "learning_rate": 6.639740701006889e-06,
+      "loss": 0.5913,
+      "step": 6465
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7000180266583456,
+      "learning_rate": 6.6387661310265315e-06,
+      "loss": 0.622,
+      "step": 6466
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6286676745558507,
+      "learning_rate": 6.637791491289136e-06,
+      "loss": 0.5497,
+      "step": 6467
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.764354709144767,
+      "learning_rate": 6.63681678183619e-06,
+      "loss": 0.5325,
+      "step": 6468
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8300264652027384,
+      "learning_rate": 6.635842002709185e-06,
+      "loss": 0.5746,
+      "step": 6469
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5822342238397752,
+      "learning_rate": 6.634867153949613e-06,
+      "loss": 0.5355,
+      "step": 6470
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6681859347938008,
+      "learning_rate": 6.633892235598971e-06,
+      "loss": 0.4999,
+      "step": 6471
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6195650881913226,
+      "learning_rate": 6.6329172476987565e-06,
+      "loss": 0.521,
+      "step": 6472
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5777079027188079,
+      "learning_rate": 6.631942190290474e-06,
+      "loss": 0.5668,
+      "step": 6473
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7482830965645356,
+      "learning_rate": 6.6309670634156265e-06,
+      "loss": 0.6085,
+      "step": 6474
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6393661927024398,
+      "learning_rate": 6.629991867115724e-06,
+      "loss": 0.5,
+      "step": 6475
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6625920334685923,
+      "learning_rate": 6.629016601432275e-06,
+      "loss": 0.4856,
+      "step": 6476
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5970890277792154,
+      "learning_rate": 6.628041266406795e-06,
+      "loss": 0.5153,
+      "step": 6477
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7767704226989249,
+      "learning_rate": 6.6270658620808015e-06,
+      "loss": 0.6396,
+      "step": 6478
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8129509103335189,
+      "learning_rate": 6.6260903884958135e-06,
+      "loss": 0.605,
+      "step": 6479
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7574390987238969,
+      "learning_rate": 6.625114845693353e-06,
+      "loss": 0.6333,
+      "step": 6480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.83783835637728,
+      "learning_rate": 6.624139233714948e-06,
+      "loss": 0.5886,
+      "step": 6481
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5744249648369125,
+      "learning_rate": 6.623163552602125e-06,
+      "loss": 0.4702,
+      "step": 6482
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6065719104766597,
+      "learning_rate": 6.622187802396416e-06,
+      "loss": 0.5621,
+      "step": 6483
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6252720927272606,
+      "learning_rate": 6.6212119831393574e-06,
+      "loss": 0.526,
+      "step": 6484
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7300166847462839,
+      "learning_rate": 6.620236094872485e-06,
+      "loss": 0.5268,
+      "step": 6485
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5193651488662671,
+      "learning_rate": 6.619260137637339e-06,
+      "loss": 0.4615,
+      "step": 6486
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6086862260498218,
+      "learning_rate": 6.618284111475464e-06,
+      "loss": 0.5297,
+      "step": 6487
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7110119872791875,
+      "learning_rate": 6.617308016428405e-06,
+      "loss": 0.614,
+      "step": 6488
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7106997797929842,
+      "learning_rate": 6.616331852537712e-06,
+      "loss": 0.5398,
+      "step": 6489
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7885781093659249,
+      "learning_rate": 6.615355619844937e-06,
+      "loss": 0.6141,
+      "step": 6490
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5541225600642373,
+      "learning_rate": 6.614379318391635e-06,
+      "loss": 0.4978,
+      "step": 6491
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.643156282920605,
+      "learning_rate": 6.613402948219365e-06,
+      "loss": 0.5638,
+      "step": 6492
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6530901853978821,
+      "learning_rate": 6.6124265093696874e-06,
+      "loss": 0.5167,
+      "step": 6493
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7271421660121189,
+      "learning_rate": 6.6114500018841655e-06,
+      "loss": 0.6107,
+      "step": 6494
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0035529231353035,
+      "learning_rate": 6.6104734258043655e-06,
+      "loss": 0.5976,
+      "step": 6495
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5941332092616888,
+      "learning_rate": 6.609496781171859e-06,
+      "loss": 0.505,
+      "step": 6496
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6774864696177553,
+      "learning_rate": 6.608520068028218e-06,
+      "loss": 0.5789,
+      "step": 6497
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6889388082929483,
+      "learning_rate": 6.607543286415017e-06,
+      "loss": 0.604,
+      "step": 6498
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0277921515906445,
+      "learning_rate": 6.606566436373836e-06,
+      "loss": 0.6065,
+      "step": 6499
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6169010073526765,
+      "learning_rate": 6.605589517946256e-06,
+      "loss": 0.5183,
+      "step": 6500
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7065624553356541,
+      "learning_rate": 6.6046125311738595e-06,
+      "loss": 0.5776,
+      "step": 6501
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8840570022135378,
+      "learning_rate": 6.603635476098236e-06,
+      "loss": 0.6505,
+      "step": 6502
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7868794113995556,
+      "learning_rate": 6.602658352760975e-06,
+      "loss": 0.6901,
+      "step": 6503
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7150974512226105,
+      "learning_rate": 6.6016811612036695e-06,
+      "loss": 0.5406,
+      "step": 6504
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5630857591089643,
+      "learning_rate": 6.600703901467914e-06,
+      "loss": 0.5951,
+      "step": 6505
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5373656574791883,
+      "learning_rate": 6.599726573595309e-06,
+      "loss": 0.4648,
+      "step": 6506
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8166571002152593,
+      "learning_rate": 6.598749177627456e-06,
+      "loss": 0.6198,
+      "step": 6507
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6807433614817046,
+      "learning_rate": 6.59777171360596e-06,
+      "loss": 0.6431,
+      "step": 6508
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7913926797743724,
+      "learning_rate": 6.596794181572428e-06,
+      "loss": 0.5567,
+      "step": 6509
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6324205471993662,
+      "learning_rate": 6.5958165815684696e-06,
+      "loss": 0.5484,
+      "step": 6510
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6253357026129646,
+      "learning_rate": 6.594838913635698e-06,
+      "loss": 0.5301,
+      "step": 6511
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5970319795024827,
+      "learning_rate": 6.5938611778157315e-06,
+      "loss": 0.533,
+      "step": 6512
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5828355647198787,
+      "learning_rate": 6.5928833741501875e-06,
+      "loss": 0.4857,
+      "step": 6513
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6678940477959378,
+      "learning_rate": 6.591905502680688e-06,
+      "loss": 0.5708,
+      "step": 6514
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.63703960912389,
+      "learning_rate": 6.590927563448859e-06,
+      "loss": 0.519,
+      "step": 6515
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7370632149374999,
+      "learning_rate": 6.5899495564963266e-06,
+      "loss": 0.6372,
+      "step": 6516
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5803613925385694,
+      "learning_rate": 6.588971481864723e-06,
+      "loss": 0.5234,
+      "step": 6517
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6550923523037336,
+      "learning_rate": 6.587993339595682e-06,
+      "loss": 0.5217,
+      "step": 6518
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7389184340573136,
+      "learning_rate": 6.587015129730839e-06,
+      "loss": 0.5622,
+      "step": 6519
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8541712774022205,
+      "learning_rate": 6.5860368523118305e-06,
+      "loss": 0.6901,
+      "step": 6520
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6462953914893367,
+      "learning_rate": 6.585058507380303e-06,
+      "loss": 0.5339,
+      "step": 6521
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7844240914151441,
+      "learning_rate": 6.584080094977901e-06,
+      "loss": 0.6381,
+      "step": 6522
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8928868013268039,
+      "learning_rate": 6.58310161514627e-06,
+      "loss": 0.6651,
+      "step": 6523
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8344502485279831,
+      "learning_rate": 6.582123067927062e-06,
+      "loss": 0.6399,
+      "step": 6524
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6902543562148888,
+      "learning_rate": 6.581144453361932e-06,
+      "loss": 0.5723,
+      "step": 6525
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7606838378903613,
+      "learning_rate": 6.580165771492535e-06,
+      "loss": 0.6172,
+      "step": 6526
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7680349264800025,
+      "learning_rate": 6.57918702236053e-06,
+      "loss": 0.5181,
+      "step": 6527
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.681643533752028,
+      "learning_rate": 6.5782082060075796e-06,
+      "loss": 0.5406,
+      "step": 6528
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6649957214325597,
+      "learning_rate": 6.5772293224753496e-06,
+      "loss": 0.5044,
+      "step": 6529
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.66792564259355,
+      "learning_rate": 6.576250371805507e-06,
+      "loss": 0.5156,
+      "step": 6530
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6299622261759319,
+      "learning_rate": 6.575271354039724e-06,
+      "loss": 0.5279,
+      "step": 6531
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.881512906590141,
+      "learning_rate": 6.5742922692196734e-06,
+      "loss": 0.6224,
+      "step": 6532
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6564445611536593,
+      "learning_rate": 6.57331311738703e-06,
+      "loss": 0.5309,
+      "step": 6533
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8131860687793555,
+      "learning_rate": 6.572333898583476e-06,
+      "loss": 0.6158,
+      "step": 6534
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6407124194578098,
+      "learning_rate": 6.571354612850693e-06,
+      "loss": 0.5341,
+      "step": 6535
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7328902208857833,
+      "learning_rate": 6.570375260230364e-06,
+      "loss": 0.5881,
+      "step": 6536
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8464478431660932,
+      "learning_rate": 6.56939584076418e-06,
+      "loss": 0.6299,
+      "step": 6537
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7191489096232528,
+      "learning_rate": 6.568416354493827e-06,
+      "loss": 0.5755,
+      "step": 6538
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7365817050537748,
+      "learning_rate": 6.5674368014610044e-06,
+      "loss": 0.5944,
+      "step": 6539
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0455182794406508,
+      "learning_rate": 6.5664571817074065e-06,
+      "loss": 0.6279,
+      "step": 6540
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6900667703590811,
+      "learning_rate": 6.56547749527473e-06,
+      "loss": 0.5508,
+      "step": 6541
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7366802134564446,
+      "learning_rate": 6.5644977422046805e-06,
+      "loss": 0.6097,
+      "step": 6542
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5854236989657675,
+      "learning_rate": 6.5635179225389615e-06,
+      "loss": 0.52,
+      "step": 6543
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7334019559981301,
+      "learning_rate": 6.5625380363192795e-06,
+      "loss": 0.5177,
+      "step": 6544
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8178026265013155,
+      "learning_rate": 6.5615580835873476e-06,
+      "loss": 0.6622,
+      "step": 6545
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.082534676157764,
+      "learning_rate": 6.560578064384879e-06,
+      "loss": 0.5514,
+      "step": 6546
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6947063486131437,
+      "learning_rate": 6.5595979787535865e-06,
+      "loss": 0.6005,
+      "step": 6547
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.716733835205423,
+      "learning_rate": 6.558617826735194e-06,
+      "loss": 0.612,
+      "step": 6548
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7176508683791621,
+      "learning_rate": 6.557637608371421e-06,
+      "loss": 0.5973,
+      "step": 6549
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6078637995408926,
+      "learning_rate": 6.5566573237039925e-06,
+      "loss": 0.5557,
+      "step": 6550
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7333544028476878,
+      "learning_rate": 6.555676972774634e-06,
+      "loss": 0.6464,
+      "step": 6551
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.562588182862289,
+      "learning_rate": 6.554696555625079e-06,
+      "loss": 0.4827,
+      "step": 6552
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8022970561682957,
+      "learning_rate": 6.553716072297061e-06,
+      "loss": 0.6269,
+      "step": 6553
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7294224063774781,
+      "learning_rate": 6.552735522832314e-06,
+      "loss": 0.6535,
+      "step": 6554
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.607388750049785,
+      "learning_rate": 6.551754907272578e-06,
+      "loss": 0.5564,
+      "step": 6555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6561026084748848,
+      "learning_rate": 6.550774225659594e-06,
+      "loss": 0.517,
+      "step": 6556
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5607195291470843,
+      "learning_rate": 6.549793478035107e-06,
+      "loss": 0.4973,
+      "step": 6557
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7643485170954416,
+      "learning_rate": 6.548812664440864e-06,
+      "loss": 0.6615,
+      "step": 6558
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5919573705467266,
+      "learning_rate": 6.547831784918614e-06,
+      "loss": 0.4944,
+      "step": 6559
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5682581819493193,
+      "learning_rate": 6.546850839510113e-06,
+      "loss": 0.5503,
+      "step": 6560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6050789187504977,
+      "learning_rate": 6.545869828257114e-06,
+      "loss": 0.5004,
+      "step": 6561
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7396563529524641,
+      "learning_rate": 6.5448887512013755e-06,
+      "loss": 0.6026,
+      "step": 6562
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7393027655715211,
+      "learning_rate": 6.5439076083846595e-06,
+      "loss": 0.6695,
+      "step": 6563
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6957350790728904,
+      "learning_rate": 6.542926399848732e-06,
+      "loss": 0.5995,
+      "step": 6564
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5852414204167266,
+      "learning_rate": 6.541945125635357e-06,
+      "loss": 0.5134,
+      "step": 6565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8122329245922367,
+      "learning_rate": 6.5409637857863066e-06,
+      "loss": 0.6121,
+      "step": 6566
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6456744614721865,
+      "learning_rate": 6.539982380343351e-06,
+      "loss": 0.5703,
+      "step": 6567
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6256095112322341,
+      "learning_rate": 6.539000909348266e-06,
+      "loss": 0.4991,
+      "step": 6568
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.661420890309588,
+      "learning_rate": 6.5380193728428294e-06,
+      "loss": 0.4821,
+      "step": 6569
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6184074991915045,
+      "learning_rate": 6.537037770868825e-06,
+      "loss": 0.6062,
+      "step": 6570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6914185402351574,
+      "learning_rate": 6.536056103468034e-06,
+      "loss": 0.5737,
+      "step": 6571
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7803554062738187,
+      "learning_rate": 6.535074370682243e-06,
+      "loss": 0.6205,
+      "step": 6572
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7003952693039684,
+      "learning_rate": 6.534092572553241e-06,
+      "loss": 0.5838,
+      "step": 6573
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6554043794474748,
+      "learning_rate": 6.533110709122821e-06,
+      "loss": 0.5549,
+      "step": 6574
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.757956701525781,
+      "learning_rate": 6.532128780432776e-06,
+      "loss": 0.5991,
+      "step": 6575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.621540951032334,
+      "learning_rate": 6.5311467865249064e-06,
+      "loss": 0.5298,
+      "step": 6576
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6822587431034667,
+      "learning_rate": 6.5301647274410105e-06,
+      "loss": 0.5544,
+      "step": 6577
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0023514807622687,
+      "learning_rate": 6.529182603222892e-06,
+      "loss": 0.5842,
+      "step": 6578
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5573906384691116,
+      "learning_rate": 6.528200413912357e-06,
+      "loss": 0.4875,
+      "step": 6579
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7442181351488076,
+      "learning_rate": 6.527218159551213e-06,
+      "loss": 0.6373,
+      "step": 6580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.711695797430948,
+      "learning_rate": 6.526235840181272e-06,
+      "loss": 0.5052,
+      "step": 6581
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6432289557796902,
+      "learning_rate": 6.525253455844348e-06,
+      "loss": 0.5628,
+      "step": 6582
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6137086632156515,
+      "learning_rate": 6.524271006582259e-06,
+      "loss": 0.5217,
+      "step": 6583
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5704356839424467,
+      "learning_rate": 6.523288492436826e-06,
+      "loss": 0.5548,
+      "step": 6584
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6205227296711358,
+      "learning_rate": 6.522305913449867e-06,
+      "loss": 0.5726,
+      "step": 6585
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7307209198799085,
+      "learning_rate": 6.521323269663211e-06,
+      "loss": 0.5787,
+      "step": 6586
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5419852748657038,
+      "learning_rate": 6.5203405611186855e-06,
+      "loss": 0.4894,
+      "step": 6587
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7233683517376814,
+      "learning_rate": 6.519357787858119e-06,
+      "loss": 0.6103,
+      "step": 6588
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6430593217668235,
+      "learning_rate": 6.518374949923347e-06,
+      "loss": 0.5225,
+      "step": 6589
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7916466069694474,
+      "learning_rate": 6.517392047356205e-06,
+      "loss": 0.5987,
+      "step": 6590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2594116491866636,
+      "learning_rate": 6.516409080198535e-06,
+      "loss": 0.6445,
+      "step": 6591
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5840893942174273,
+      "learning_rate": 6.5154260484921735e-06,
+      "loss": 0.5139,
+      "step": 6592
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7183332341164766,
+      "learning_rate": 6.51444295227897e-06,
+      "loss": 0.5961,
+      "step": 6593
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8634584307887804,
+      "learning_rate": 6.513459791600768e-06,
+      "loss": 0.6501,
+      "step": 6594
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7755051665441314,
+      "learning_rate": 6.51247656649942e-06,
+      "loss": 0.6365,
+      "step": 6595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.19429483664414,
+      "learning_rate": 6.511493277016776e-06,
+      "loss": 0.574,
+      "step": 6596
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6991586381722998,
+      "learning_rate": 6.510509923194694e-06,
+      "loss": 0.56,
+      "step": 6597
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5917032495669147,
+      "learning_rate": 6.509526505075032e-06,
+      "loss": 0.5403,
+      "step": 6598
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6509379503007197,
+      "learning_rate": 6.508543022699652e-06,
+      "loss": 0.6081,
+      "step": 6599
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6189794787068471,
+      "learning_rate": 6.5075594761104126e-06,
+      "loss": 0.5519,
+      "step": 6600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6791622848975046,
+      "learning_rate": 6.506575865349185e-06,
+      "loss": 0.5481,
+      "step": 6601
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7531039358858942,
+      "learning_rate": 6.505592190457838e-06,
+      "loss": 0.6141,
+      "step": 6602
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8505550260254032,
+      "learning_rate": 6.504608451478242e-06,
+      "loss": 0.5806,
+      "step": 6603
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0209036518929284,
+      "learning_rate": 6.503624648452273e-06,
+      "loss": 0.6671,
+      "step": 6604
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6238119148373528,
+      "learning_rate": 6.502640781421807e-06,
+      "loss": 0.5509,
+      "step": 6605
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0555480266865047,
+      "learning_rate": 6.501656850428724e-06,
+      "loss": 0.6019,
+      "step": 6606
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5895111872159797,
+      "learning_rate": 6.500672855514909e-06,
+      "loss": 0.5206,
+      "step": 6607
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7822501077831937,
+      "learning_rate": 6.499688796722244e-06,
+      "loss": 0.6117,
+      "step": 6608
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6030800136079699,
+      "learning_rate": 6.498704674092622e-06,
+      "loss": 0.4907,
+      "step": 6609
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7311892330290485,
+      "learning_rate": 6.497720487667929e-06,
+      "loss": 0.6285,
+      "step": 6610
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6095372510185418,
+      "learning_rate": 6.496736237490062e-06,
+      "loss": 0.5183,
+      "step": 6611
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6074076574843538,
+      "learning_rate": 6.4957519236009156e-06,
+      "loss": 0.5039,
+      "step": 6612
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7935921019013585,
+      "learning_rate": 6.494767546042389e-06,
+      "loss": 0.5032,
+      "step": 6613
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7714659687510147,
+      "learning_rate": 6.493783104856386e-06,
+      "loss": 0.6196,
+      "step": 6614
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5924889971866503,
+      "learning_rate": 6.492798600084809e-06,
+      "loss": 0.5081,
+      "step": 6615
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5687662927402892,
+      "learning_rate": 6.491814031769566e-06,
+      "loss": 0.4873,
+      "step": 6616
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7601024218974984,
+      "learning_rate": 6.4908293999525675e-06,
+      "loss": 0.6131,
+      "step": 6617
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1927784118332903,
+      "learning_rate": 6.489844704675724e-06,
+      "loss": 0.6214,
+      "step": 6618
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7791536561609305,
+      "learning_rate": 6.488859945980952e-06,
+      "loss": 0.6187,
+      "step": 6619
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6684148761859198,
+      "learning_rate": 6.4878751239101715e-06,
+      "loss": 0.5644,
+      "step": 6620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7172590504768079,
+      "learning_rate": 6.486890238505301e-06,
+      "loss": 0.5982,
+      "step": 6621
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6198513943762066,
+      "learning_rate": 6.485905289808264e-06,
+      "loss": 0.5062,
+      "step": 6622
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6007609126335464,
+      "learning_rate": 6.484920277860988e-06,
+      "loss": 0.5225,
+      "step": 6623
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6454183493838301,
+      "learning_rate": 6.4839352027054e-06,
+      "loss": 0.519,
+      "step": 6624
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7218154042427212,
+      "learning_rate": 6.482950064383432e-06,
+      "loss": 0.5352,
+      "step": 6625
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7128891104835277,
+      "learning_rate": 6.481964862937019e-06,
+      "loss": 0.5917,
+      "step": 6626
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6481267742263872,
+      "learning_rate": 6.480979598408098e-06,
+      "loss": 0.4986,
+      "step": 6627
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.576561963314315,
+      "learning_rate": 6.479994270838608e-06,
+      "loss": 0.5241,
+      "step": 6628
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7723381342276889,
+      "learning_rate": 6.479008880270491e-06,
+      "loss": 0.6232,
+      "step": 6629
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5712802892892588,
+      "learning_rate": 6.478023426745692e-06,
+      "loss": 0.5356,
+      "step": 6630
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6218614701393628,
+      "learning_rate": 6.477037910306159e-06,
+      "loss": 0.4964,
+      "step": 6631
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7774841211352943,
+      "learning_rate": 6.476052330993842e-06,
+      "loss": 0.6049,
+      "step": 6632
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5631581229202345,
+      "learning_rate": 6.475066688850694e-06,
+      "loss": 0.4696,
+      "step": 6633
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5174789809951127,
+      "learning_rate": 6.474080983918671e-06,
+      "loss": 0.4532,
+      "step": 6634
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6606388601648999,
+      "learning_rate": 6.473095216239731e-06,
+      "loss": 0.5051,
+      "step": 6635
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8533707933848041,
+      "learning_rate": 6.472109385855836e-06,
+      "loss": 0.6202,
+      "step": 6636
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6558808481903375,
+      "learning_rate": 6.471123492808946e-06,
+      "loss": 0.4936,
+      "step": 6637
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7955505713043879,
+      "learning_rate": 6.470137537141031e-06,
+      "loss": 0.6064,
+      "step": 6638
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7829917928962898,
+      "learning_rate": 6.46915151889406e-06,
+      "loss": 0.6635,
+      "step": 6639
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.770157181848176,
+      "learning_rate": 6.468165438110004e-06,
+      "loss": 0.5681,
+      "step": 6640
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6250488830721793,
+      "learning_rate": 6.467179294830835e-06,
+      "loss": 0.5073,
+      "step": 6641
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6868334415337548,
+      "learning_rate": 6.4661930890985335e-06,
+      "loss": 0.5617,
+      "step": 6642
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7920069716883784,
+      "learning_rate": 6.465206820955076e-06,
+      "loss": 0.6303,
+      "step": 6643
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.715335768436534,
+      "learning_rate": 6.464220490442446e-06,
+      "loss": 0.6082,
+      "step": 6644
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7336124810956771,
+      "learning_rate": 6.463234097602631e-06,
+      "loss": 0.4735,
+      "step": 6645
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7202239309172377,
+      "learning_rate": 6.462247642477615e-06,
+      "loss": 0.545,
+      "step": 6646
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5941892673870489,
+      "learning_rate": 6.461261125109389e-06,
+      "loss": 0.5382,
+      "step": 6647
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5826192815089393,
+      "learning_rate": 6.460274545539947e-06,
+      "loss": 0.5037,
+      "step": 6648
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8123260487299397,
+      "learning_rate": 6.459287903811285e-06,
+      "loss": 0.5952,
+      "step": 6649
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9380444776394964,
+      "learning_rate": 6.458301199965398e-06,
+      "loss": 0.6765,
+      "step": 6650
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5444336617343107,
+      "learning_rate": 6.457314434044289e-06,
+      "loss": 0.5142,
+      "step": 6651
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0043128034882705,
+      "learning_rate": 6.456327606089962e-06,
+      "loss": 0.6496,
+      "step": 6652
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8263048961454296,
+      "learning_rate": 6.455340716144425e-06,
+      "loss": 0.6103,
+      "step": 6653
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5912204655334908,
+      "learning_rate": 6.4543537642496816e-06,
+      "loss": 0.5506,
+      "step": 6654
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5919250552538146,
+      "learning_rate": 6.453366750447747e-06,
+      "loss": 0.545,
+      "step": 6655
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5435661626258214,
+      "learning_rate": 6.452379674780633e-06,
+      "loss": 0.5013,
+      "step": 6656
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7510925673391442,
+      "learning_rate": 6.45139253729036e-06,
+      "loss": 0.5604,
+      "step": 6657
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8422280437779414,
+      "learning_rate": 6.450405338018943e-06,
+      "loss": 0.689,
+      "step": 6658
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7140162987163611,
+      "learning_rate": 6.449418077008405e-06,
+      "loss": 0.6538,
+      "step": 6659
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5467812951779568,
+      "learning_rate": 6.448430754300772e-06,
+      "loss": 0.4883,
+      "step": 6660
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.714976472578544,
+      "learning_rate": 6.44744336993807e-06,
+      "loss": 0.5659,
+      "step": 6661
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8578209927853342,
+      "learning_rate": 6.446455923962328e-06,
+      "loss": 0.5807,
+      "step": 6662
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7644429897205445,
+      "learning_rate": 6.44546841641558e-06,
+      "loss": 0.5787,
+      "step": 6663
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6509814584879079,
+      "learning_rate": 6.444480847339862e-06,
+      "loss": 0.5612,
+      "step": 6664
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.599574322432731,
+      "learning_rate": 6.443493216777209e-06,
+      "loss": 0.539,
+      "step": 6665
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6536085707771628,
+      "learning_rate": 6.442505524769663e-06,
+      "loss": 0.548,
+      "step": 6666
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8134178203412349,
+      "learning_rate": 6.441517771359267e-06,
+      "loss": 0.6095,
+      "step": 6667
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6252589088357704,
+      "learning_rate": 6.440529956588064e-06,
+      "loss": 0.4954,
+      "step": 6668
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5902260662712403,
+      "learning_rate": 6.439542080498105e-06,
+      "loss": 0.5406,
+      "step": 6669
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.660871271968508,
+      "learning_rate": 6.43855414313144e-06,
+      "loss": 0.5505,
+      "step": 6670
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6412905588827149,
+      "learning_rate": 6.437566144530123e-06,
+      "loss": 0.5753,
+      "step": 6671
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.504343457256404,
+      "learning_rate": 6.436578084736207e-06,
+      "loss": 0.4828,
+      "step": 6672
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6024850680333286,
+      "learning_rate": 6.435589963791754e-06,
+      "loss": 0.5062,
+      "step": 6673
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7674430865032238,
+      "learning_rate": 6.434601781738824e-06,
+      "loss": 0.5023,
+      "step": 6674
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8091307097736599,
+      "learning_rate": 6.433613538619479e-06,
+      "loss": 0.5952,
+      "step": 6675
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7935332083526491,
+      "learning_rate": 6.432625234475788e-06,
+      "loss": 0.6256,
+      "step": 6676
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8335983934485832,
+      "learning_rate": 6.431636869349818e-06,
+      "loss": 0.6417,
+      "step": 6677
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7133436863802591,
+      "learning_rate": 6.430648443283642e-06,
+      "loss": 0.6089,
+      "step": 6678
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7133959633483786,
+      "learning_rate": 6.4296599563193344e-06,
+      "loss": 0.6138,
+      "step": 6679
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.825994999486831,
+      "learning_rate": 6.428671408498969e-06,
+      "loss": 0.5563,
+      "step": 6680
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6029731726032074,
+      "learning_rate": 6.427682799864628e-06,
+      "loss": 0.5331,
+      "step": 6681
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6042192317360487,
+      "learning_rate": 6.426694130458392e-06,
+      "loss": 0.5252,
+      "step": 6682
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6103765821083451,
+      "learning_rate": 6.425705400322346e-06,
+      "loss": 0.5342,
+      "step": 6683
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.712333074987951,
+      "learning_rate": 6.424716609498578e-06,
+      "loss": 0.6003,
+      "step": 6684
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7858545673161307,
+      "learning_rate": 6.423727758029175e-06,
+      "loss": 0.5751,
+      "step": 6685
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6271358504217927,
+      "learning_rate": 6.422738845956232e-06,
+      "loss": 0.5103,
+      "step": 6686
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6542777104160439,
+      "learning_rate": 6.421749873321843e-06,
+      "loss": 0.5235,
+      "step": 6687
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8814752299022276,
+      "learning_rate": 6.4207608401681045e-06,
+      "loss": 0.5681,
+      "step": 6688
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6455099468820157,
+      "learning_rate": 6.419771746537118e-06,
+      "loss": 0.5141,
+      "step": 6689
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7389888470521515,
+      "learning_rate": 6.418782592470985e-06,
+      "loss": 0.584,
+      "step": 6690
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6119270179492299,
+      "learning_rate": 6.4177933780118094e-06,
+      "loss": 0.5489,
+      "step": 6691
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8741890718063151,
+      "learning_rate": 6.4168041032017026e-06,
+      "loss": 0.6322,
+      "step": 6692
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6098622082197017,
+      "learning_rate": 6.41581476808277e-06,
+      "loss": 0.5259,
+      "step": 6693
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.780727278145902,
+      "learning_rate": 6.4148253726971286e-06,
+      "loss": 0.618,
+      "step": 6694
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6646163861176623,
+      "learning_rate": 6.413835917086892e-06,
+      "loss": 0.5742,
+      "step": 6695
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7168848602542783,
+      "learning_rate": 6.4128464012941795e-06,
+      "loss": 0.5396,
+      "step": 6696
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5705818002988751,
+      "learning_rate": 6.411856825361109e-06,
+      "loss": 0.5351,
+      "step": 6697
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7337958905809268,
+      "learning_rate": 6.410867189329806e-06,
+      "loss": 0.5732,
+      "step": 6698
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1460587253276808,
+      "learning_rate": 6.4098774932423935e-06,
+      "loss": 0.5923,
+      "step": 6699
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7969143690271436,
+      "learning_rate": 6.408887737141003e-06,
+      "loss": 0.598,
+      "step": 6700
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6012642235502653,
+      "learning_rate": 6.407897921067763e-06,
+      "loss": 0.5133,
+      "step": 6701
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6241253434314572,
+      "learning_rate": 6.406908045064808e-06,
+      "loss": 0.5352,
+      "step": 6702
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.603500412991251,
+      "learning_rate": 6.405918109174274e-06,
+      "loss": 0.5635,
+      "step": 6703
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7130743433358312,
+      "learning_rate": 6.404928113438298e-06,
+      "loss": 0.6062,
+      "step": 6704
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8390679425955562,
+      "learning_rate": 6.403938057899021e-06,
+      "loss": 0.5896,
+      "step": 6705
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6003620047611676,
+      "learning_rate": 6.402947942598588e-06,
+      "loss": 0.4483,
+      "step": 6706
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7809477775062855,
+      "learning_rate": 6.401957767579143e-06,
+      "loss": 0.6145,
+      "step": 6707
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7214247044322019,
+      "learning_rate": 6.400967532882838e-06,
+      "loss": 0.5197,
+      "step": 6708
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0330520167490873,
+      "learning_rate": 6.399977238551821e-06,
+      "loss": 0.6049,
+      "step": 6709
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.538280732392651,
+      "learning_rate": 6.398986884628246e-06,
+      "loss": 0.5185,
+      "step": 6710
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7801000993264211,
+      "learning_rate": 6.397996471154272e-06,
+      "loss": 0.6316,
+      "step": 6711
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6546423021000509,
+      "learning_rate": 6.3970059981720526e-06,
+      "loss": 0.5182,
+      "step": 6712
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8467742519251735,
+      "learning_rate": 6.396015465723754e-06,
+      "loss": 0.6319,
+      "step": 6713
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6799915195617798,
+      "learning_rate": 6.395024873851537e-06,
+      "loss": 0.5505,
+      "step": 6714
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7187111517934552,
+      "learning_rate": 6.394034222597571e-06,
+      "loss": 0.5635,
+      "step": 6715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6112867117847852,
+      "learning_rate": 6.393043512004022e-06,
+      "loss": 0.5009,
+      "step": 6716
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7582173584268032,
+      "learning_rate": 6.39205274211306e-06,
+      "loss": 0.6385,
+      "step": 6717
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6696693004498872,
+      "learning_rate": 6.391061912966864e-06,
+      "loss": 0.5543,
+      "step": 6718
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6591884404022977,
+      "learning_rate": 6.3900710246076055e-06,
+      "loss": 0.5493,
+      "step": 6719
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7622445422933825,
+      "learning_rate": 6.389080077077467e-06,
+      "loss": 0.6071,
+      "step": 6720
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7749665447037803,
+      "learning_rate": 6.3880890704186285e-06,
+      "loss": 0.6113,
+      "step": 6721
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5555647932447542,
+      "learning_rate": 6.387098004673274e-06,
+      "loss": 0.4722,
+      "step": 6722
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.591486054805839,
+      "learning_rate": 6.386106879883589e-06,
+      "loss": 0.5137,
+      "step": 6723
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6986612083317825,
+      "learning_rate": 6.385115696091763e-06,
+      "loss": 0.5689,
+      "step": 6724
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6413008031795211,
+      "learning_rate": 6.384124453339988e-06,
+      "loss": 0.5151,
+      "step": 6725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6807008739889637,
+      "learning_rate": 6.38313315167046e-06,
+      "loss": 0.5474,
+      "step": 6726
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7190364538883304,
+      "learning_rate": 6.382141791125371e-06,
+      "loss": 0.6143,
+      "step": 6727
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7138245656461645,
+      "learning_rate": 6.381150371746925e-06,
+      "loss": 0.5144,
+      "step": 6728
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6731068278629442,
+      "learning_rate": 6.380158893577321e-06,
+      "loss": 0.539,
+      "step": 6729
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6334266464840334,
+      "learning_rate": 6.379167356658761e-06,
+      "loss": 0.4887,
+      "step": 6730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6136014123301498,
+      "learning_rate": 6.378175761033456e-06,
+      "loss": 0.5422,
+      "step": 6731
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5998200679599752,
+      "learning_rate": 6.377184106743612e-06,
+      "loss": 0.5262,
+      "step": 6732
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5659747295383475,
+      "learning_rate": 6.3761923938314415e-06,
+      "loss": 0.5157,
+      "step": 6733
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5649108934638767,
+      "learning_rate": 6.375200622339159e-06,
+      "loss": 0.4787,
+      "step": 6734
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7574165968262838,
+      "learning_rate": 6.37420879230898e-06,
+      "loss": 0.5744,
+      "step": 6735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4388942882369706,
+      "learning_rate": 6.3732169037831246e-06,
+      "loss": 0.6236,
+      "step": 6736
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9795215249888689,
+      "learning_rate": 6.372224956803812e-06,
+      "loss": 0.622,
+      "step": 6737
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5670350333469152,
+      "learning_rate": 6.37123295141327e-06,
+      "loss": 0.5263,
+      "step": 6738
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6845583822661014,
+      "learning_rate": 6.3702408876537224e-06,
+      "loss": 0.5792,
+      "step": 6739
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6381552008092994,
+      "learning_rate": 6.3692487655674e-06,
+      "loss": 0.5252,
+      "step": 6740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5986972026279694,
+      "learning_rate": 6.368256585196532e-06,
+      "loss": 0.552,
+      "step": 6741
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7205224659588964,
+      "learning_rate": 6.3672643465833525e-06,
+      "loss": 0.6018,
+      "step": 6742
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8062345365126261,
+      "learning_rate": 6.3662720497700996e-06,
+      "loss": 0.5429,
+      "step": 6743
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6414281985888026,
+      "learning_rate": 6.365279694799012e-06,
+      "loss": 0.5147,
+      "step": 6744
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7556485437413981,
+      "learning_rate": 6.364287281712331e-06,
+      "loss": 0.5875,
+      "step": 6745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5826631493802104,
+      "learning_rate": 6.3632948105522995e-06,
+      "loss": 0.4515,
+      "step": 6746
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7531103371860258,
+      "learning_rate": 6.362302281361165e-06,
+      "loss": 0.6538,
+      "step": 6747
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7466964161900267,
+      "learning_rate": 6.361309694181175e-06,
+      "loss": 0.6341,
+      "step": 6748
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5391393020601636,
+      "learning_rate": 6.3603170490545805e-06,
+      "loss": 0.5266,
+      "step": 6749
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7314106997727251,
+      "learning_rate": 6.3593243460236376e-06,
+      "loss": 0.5822,
+      "step": 6750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7067591723768756,
+      "learning_rate": 6.3583315851306e-06,
+      "loss": 0.571,
+      "step": 6751
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5537794363673365,
+      "learning_rate": 6.357338766417729e-06,
+      "loss": 0.5056,
+      "step": 6752
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6287934355885714,
+      "learning_rate": 6.356345889927284e-06,
+      "loss": 0.5547,
+      "step": 6753
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7113761522520146,
+      "learning_rate": 6.355352955701528e-06,
+      "loss": 0.6214,
+      "step": 6754
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9750561780835462,
+      "learning_rate": 6.354359963782726e-06,
+      "loss": 0.619,
+      "step": 6755
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8919735503043947,
+      "learning_rate": 6.353366914213151e-06,
+      "loss": 0.6053,
+      "step": 6756
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.709234512999577,
+      "learning_rate": 6.3523738070350705e-06,
+      "loss": 0.6305,
+      "step": 6757
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8205006112398691,
+      "learning_rate": 6.351380642290757e-06,
+      "loss": 0.6353,
+      "step": 6758
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5574928899055271,
+      "learning_rate": 6.35038742002249e-06,
+      "loss": 0.4891,
+      "step": 6759
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5304699646174911,
+      "learning_rate": 6.349394140272546e-06,
+      "loss": 0.5112,
+      "step": 6760
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7501973588898401,
+      "learning_rate": 6.348400803083204e-06,
+      "loss": 0.5445,
+      "step": 6761
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9388017352297809,
+      "learning_rate": 6.347407408496749e-06,
+      "loss": 0.5875,
+      "step": 6762
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7173405857868247,
+      "learning_rate": 6.346413956555466e-06,
+      "loss": 0.5842,
+      "step": 6763
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8187707723663016,
+      "learning_rate": 6.345420447301644e-06,
+      "loss": 0.6533,
+      "step": 6764
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7417909680826302,
+      "learning_rate": 6.3444268807775736e-06,
+      "loss": 0.5896,
+      "step": 6765
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7642172112967541,
+      "learning_rate": 6.3434332570255465e-06,
+      "loss": 0.6128,
+      "step": 6766
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6114815433787782,
+      "learning_rate": 6.342439576087858e-06,
+      "loss": 0.5052,
+      "step": 6767
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0236714350210854,
+      "learning_rate": 6.341445838006806e-06,
+      "loss": 0.5985,
+      "step": 6768
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6744256826388978,
+      "learning_rate": 6.340452042824693e-06,
+      "loss": 0.4624,
+      "step": 6769
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7222292963664174,
+      "learning_rate": 6.339458190583819e-06,
+      "loss": 0.5547,
+      "step": 6770
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7537838566144314,
+      "learning_rate": 6.3384642813264904e-06,
+      "loss": 0.6251,
+      "step": 6771
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5984447339557778,
+      "learning_rate": 6.337470315095016e-06,
+      "loss": 0.5484,
+      "step": 6772
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5093405123693988,
+      "learning_rate": 6.336476291931702e-06,
+      "loss": 0.4885,
+      "step": 6773
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6065780791658569,
+      "learning_rate": 6.3354822118788624e-06,
+      "loss": 0.5449,
+      "step": 6774
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7501777419214004,
+      "learning_rate": 6.334488074978815e-06,
+      "loss": 0.5172,
+      "step": 6775
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7495723344250118,
+      "learning_rate": 6.3334938812738734e-06,
+      "loss": 0.5974,
+      "step": 6776
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7150863817269848,
+      "learning_rate": 6.332499630806359e-06,
+      "loss": 0.5837,
+      "step": 6777
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7851310025120265,
+      "learning_rate": 6.3315053236185935e-06,
+      "loss": 0.6754,
+      "step": 6778
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7222751470063057,
+      "learning_rate": 6.330510959752902e-06,
+      "loss": 0.6121,
+      "step": 6779
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9211548415234667,
+      "learning_rate": 6.329516539251609e-06,
+      "loss": 0.5665,
+      "step": 6780
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5627713639444607,
+      "learning_rate": 6.328522062157045e-06,
+      "loss": 0.5454,
+      "step": 6781
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8162151870739535,
+      "learning_rate": 6.327527528511544e-06,
+      "loss": 0.586,
+      "step": 6782
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6706600262678507,
+      "learning_rate": 6.326532938357438e-06,
+      "loss": 0.554,
+      "step": 6783
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7132656147409749,
+      "learning_rate": 6.3255382917370634e-06,
+      "loss": 0.6419,
+      "step": 6784
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5587804142270424,
+      "learning_rate": 6.32454358869276e-06,
+      "loss": 0.5254,
+      "step": 6785
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5622653100116567,
+      "learning_rate": 6.323548829266867e-06,
+      "loss": 0.5008,
+      "step": 6786
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.638376379275469,
+      "learning_rate": 6.322554013501731e-06,
+      "loss": 0.5835,
+      "step": 6787
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6609193734697818,
+      "learning_rate": 6.321559141439697e-06,
+      "loss": 0.5525,
+      "step": 6788
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6013839173079497,
+      "learning_rate": 6.320564213123113e-06,
+      "loss": 0.563,
+      "step": 6789
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6008060519934829,
+      "learning_rate": 6.319569228594331e-06,
+      "loss": 0.4826,
+      "step": 6790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6721232126266182,
+      "learning_rate": 6.318574187895703e-06,
+      "loss": 0.5577,
+      "step": 6791
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.532686461543334,
+      "learning_rate": 6.3175790910695845e-06,
+      "loss": 0.5214,
+      "step": 6792
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5937242019250427,
+      "learning_rate": 6.316583938158336e-06,
+      "loss": 0.4935,
+      "step": 6793
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.65146126396118,
+      "learning_rate": 6.3155887292043164e-06,
+      "loss": 0.495,
+      "step": 6794
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7826559495731986,
+      "learning_rate": 6.314593464249889e-06,
+      "loss": 0.5865,
+      "step": 6795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6054753434688831,
+      "learning_rate": 6.313598143337417e-06,
+      "loss": 0.5177,
+      "step": 6796
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7828603925191665,
+      "learning_rate": 6.312602766509271e-06,
+      "loss": 0.5594,
+      "step": 6797
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5976344886995849,
+      "learning_rate": 6.31160733380782e-06,
+      "loss": 0.5399,
+      "step": 6798
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7608514391926767,
+      "learning_rate": 6.310611845275434e-06,
+      "loss": 0.5393,
+      "step": 6799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7107898596213578,
+      "learning_rate": 6.309616300954492e-06,
+      "loss": 0.6108,
+      "step": 6800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.608412541856247,
+      "learning_rate": 6.308620700887368e-06,
+      "loss": 0.5478,
+      "step": 6801
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6038700595244411,
+      "learning_rate": 6.307625045116443e-06,
+      "loss": 0.5509,
+      "step": 6802
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.543431507757396,
+      "learning_rate": 6.306629333684099e-06,
+      "loss": 0.6488,
+      "step": 6803
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6243108845268378,
+      "learning_rate": 6.305633566632719e-06,
+      "loss": 0.5401,
+      "step": 6804
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8794502147054255,
+      "learning_rate": 6.3046377440046895e-06,
+      "loss": 0.6461,
+      "step": 6805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5867167302657583,
+      "learning_rate": 6.3036418658424e-06,
+      "loss": 0.5354,
+      "step": 6806
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6031382569390703,
+      "learning_rate": 6.3026459321882435e-06,
+      "loss": 0.5559,
+      "step": 6807
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5626801660294306,
+      "learning_rate": 6.301649943084612e-06,
+      "loss": 0.5391,
+      "step": 6808
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7008523751937906,
+      "learning_rate": 6.300653898573903e-06,
+      "loss": 0.5719,
+      "step": 6809
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6202315961834426,
+      "learning_rate": 6.299657798698512e-06,
+      "loss": 0.5435,
+      "step": 6810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7578252059528312,
+      "learning_rate": 6.2986616435008415e-06,
+      "loss": 0.6158,
+      "step": 6811
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6815939664744214,
+      "learning_rate": 6.297665433023295e-06,
+      "loss": 0.554,
+      "step": 6812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7057340399865087,
+      "learning_rate": 6.296669167308279e-06,
+      "loss": 0.6013,
+      "step": 6813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6153740877570159,
+      "learning_rate": 6.2956728463982e-06,
+      "loss": 0.4992,
+      "step": 6814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5921791749086263,
+      "learning_rate": 6.294676470335468e-06,
+      "loss": 0.4803,
+      "step": 6815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7411634893068696,
+      "learning_rate": 6.293680039162495e-06,
+      "loss": 0.6038,
+      "step": 6816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6545407829797351,
+      "learning_rate": 6.292683552921697e-06,
+      "loss": 0.5949,
+      "step": 6817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7276433511793893,
+      "learning_rate": 6.291687011655491e-06,
+      "loss": 0.5784,
+      "step": 6818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7790355606637095,
+      "learning_rate": 6.290690415406297e-06,
+      "loss": 0.6273,
+      "step": 6819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8857327667942153,
+      "learning_rate": 6.2896937642165354e-06,
+      "loss": 0.5735,
+      "step": 6820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6715998134174386,
+      "learning_rate": 6.2886970581286335e-06,
+      "loss": 0.5292,
+      "step": 6821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7733025018650902,
+      "learning_rate": 6.287700297185015e-06,
+      "loss": 0.6041,
+      "step": 6822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7173777103567918,
+      "learning_rate": 6.286703481428109e-06,
+      "loss": 0.6402,
+      "step": 6823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.642865696458957,
+      "learning_rate": 6.2857066109003484e-06,
+      "loss": 0.5225,
+      "step": 6824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7270170677998824,
+      "learning_rate": 6.284709685644166e-06,
+      "loss": 0.5884,
+      "step": 6825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5501905168697819,
+      "learning_rate": 6.283712705701997e-06,
+      "loss": 0.5299,
+      "step": 6826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5984007750484809,
+      "learning_rate": 6.2827156711162805e-06,
+      "loss": 0.5407,
+      "step": 6827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6576386282479347,
+      "learning_rate": 6.281718581929457e-06,
+      "loss": 0.5038,
+      "step": 6828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6924926115000757,
+      "learning_rate": 6.280721438183969e-06,
+      "loss": 0.5721,
+      "step": 6829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7452618188694167,
+      "learning_rate": 6.279724239922261e-06,
+      "loss": 0.5892,
+      "step": 6830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.671576464309829,
+      "learning_rate": 6.278726987186783e-06,
+      "loss": 0.5177,
+      "step": 6831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7949311868949315,
+      "learning_rate": 6.277729680019984e-06,
+      "loss": 0.5861,
+      "step": 6832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6807662791089962,
+      "learning_rate": 6.276732318464314e-06,
+      "loss": 0.6374,
+      "step": 6833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6565947205416856,
+      "learning_rate": 6.27573490256223e-06,
+      "loss": 0.559,
+      "step": 6834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6133906144028313,
+      "learning_rate": 6.274737432356187e-06,
+      "loss": 0.5161,
+      "step": 6835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5959387865678492,
+      "learning_rate": 6.273739907888645e-06,
+      "loss": 0.5299,
+      "step": 6836
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6739268664204052,
+      "learning_rate": 6.272742329202066e-06,
+      "loss": 0.561,
+      "step": 6837
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9652026442339495,
+      "learning_rate": 6.271744696338913e-06,
+      "loss": 0.6284,
+      "step": 6838
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6663780218655141,
+      "learning_rate": 6.270747009341652e-06,
+      "loss": 0.5483,
+      "step": 6839
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6507891057327294,
+      "learning_rate": 6.269749268252753e-06,
+      "loss": 0.5399,
+      "step": 6840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6395465662584445,
+      "learning_rate": 6.268751473114684e-06,
+      "loss": 0.4978,
+      "step": 6841
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7601071188929123,
+      "learning_rate": 6.267753623969919e-06,
+      "loss": 0.5551,
+      "step": 6842
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8001905641435049,
+      "learning_rate": 6.266755720860933e-06,
+      "loss": 0.5876,
+      "step": 6843
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7579055164094853,
+      "learning_rate": 6.2657577638302045e-06,
+      "loss": 0.5557,
+      "step": 6844
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7746683846586531,
+      "learning_rate": 6.2647597529202135e-06,
+      "loss": 0.6439,
+      "step": 6845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9527593327937378,
+      "learning_rate": 6.263761688173441e-06,
+      "loss": 0.6052,
+      "step": 6846
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5798480463925557,
+      "learning_rate": 6.262763569632371e-06,
+      "loss": 0.5342,
+      "step": 6847
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7416176034640507,
+      "learning_rate": 6.261765397339491e-06,
+      "loss": 0.5626,
+      "step": 6848
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6264930743913123,
+      "learning_rate": 6.26076717133729e-06,
+      "loss": 0.5739,
+      "step": 6849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8851478382258816,
+      "learning_rate": 6.259768891668261e-06,
+      "loss": 0.6162,
+      "step": 6850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8280457744640813,
+      "learning_rate": 6.2587705583748945e-06,
+      "loss": 0.5486,
+      "step": 6851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6002522558786202,
+      "learning_rate": 6.257772171499687e-06,
+      "loss": 0.5568,
+      "step": 6852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9150153674525723,
+      "learning_rate": 6.256773731085139e-06,
+      "loss": 0.6725,
+      "step": 6853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7827022016153788,
+      "learning_rate": 6.255775237173749e-06,
+      "loss": 0.6007,
+      "step": 6854
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5676984860950057,
+      "learning_rate": 6.25477668980802e-06,
+      "loss": 0.5371,
+      "step": 6855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6547096461774186,
+      "learning_rate": 6.2537780890304565e-06,
+      "loss": 0.5605,
+      "step": 6856
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.623009421679491,
+      "learning_rate": 6.2527794348835666e-06,
+      "loss": 0.4892,
+      "step": 6857
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.543909343147493,
+      "learning_rate": 6.251780727409861e-06,
+      "loss": 0.512,
+      "step": 6858
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6814961859316504,
+      "learning_rate": 6.250781966651847e-06,
+      "loss": 0.525,
+      "step": 6859
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7807760097938817,
+      "learning_rate": 6.249783152652045e-06,
+      "loss": 0.5737,
+      "step": 6860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7379681688085484,
+      "learning_rate": 6.248784285452964e-06,
+      "loss": 0.5696,
+      "step": 6861
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.691373783756208,
+      "learning_rate": 6.247785365097129e-06,
+      "loss": 0.5181,
+      "step": 6862
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6137904787440552,
+      "learning_rate": 6.246786391627058e-06,
+      "loss": 0.5563,
+      "step": 6863
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5588059699715074,
+      "learning_rate": 6.245787365085274e-06,
+      "loss": 0.5149,
+      "step": 6864
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7195083567222281,
+      "learning_rate": 6.244788285514304e-06,
+      "loss": 0.5248,
+      "step": 6865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6688699464853312,
+      "learning_rate": 6.2437891529566745e-06,
+      "loss": 0.6034,
+      "step": 6866
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6131091843852191,
+      "learning_rate": 6.242789967454913e-06,
+      "loss": 0.5273,
+      "step": 6867
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5423820291017064,
+      "learning_rate": 6.241790729051555e-06,
+      "loss": 0.4823,
+      "step": 6868
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.542137685061483,
+      "learning_rate": 6.2407914377891355e-06,
+      "loss": 0.5467,
+      "step": 6869
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7002896528417133,
+      "learning_rate": 6.239792093710189e-06,
+      "loss": 0.6,
+      "step": 6870
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8505814826815032,
+      "learning_rate": 6.238792696857253e-06,
+      "loss": 0.6131,
+      "step": 6871
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5473859067486834,
+      "learning_rate": 6.237793247272872e-06,
+      "loss": 0.4759,
+      "step": 6872
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.899652362697501,
+      "learning_rate": 6.236793744999587e-06,
+      "loss": 0.6738,
+      "step": 6873
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8467807264115054,
+      "learning_rate": 6.2357941900799455e-06,
+      "loss": 0.6142,
+      "step": 6874
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6985128921024746,
+      "learning_rate": 6.234794582556494e-06,
+      "loss": 0.5553,
+      "step": 6875
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7619096079649466,
+      "learning_rate": 6.233794922471783e-06,
+      "loss": 0.612,
+      "step": 6876
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6663012159165116,
+      "learning_rate": 6.232795209868365e-06,
+      "loss": 0.522,
+      "step": 6877
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7446578268848328,
+      "learning_rate": 6.231795444788794e-06,
+      "loss": 0.6283,
+      "step": 6878
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5450036146527193,
+      "learning_rate": 6.230795627275626e-06,
+      "loss": 0.5677,
+      "step": 6879
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.758724870171053,
+      "learning_rate": 6.229795757371423e-06,
+      "loss": 0.5633,
+      "step": 6880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8175139850721489,
+      "learning_rate": 6.228795835118744e-06,
+      "loss": 0.6056,
+      "step": 6881
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6643489122575102,
+      "learning_rate": 6.227795860560153e-06,
+      "loss": 0.5483,
+      "step": 6882
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7445691382149575,
+      "learning_rate": 6.226795833738216e-06,
+      "loss": 0.6273,
+      "step": 6883
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5875772040990337,
+      "learning_rate": 6.225795754695501e-06,
+      "loss": 0.5116,
+      "step": 6884
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9325925337166909,
+      "learning_rate": 6.224795623474576e-06,
+      "loss": 0.6087,
+      "step": 6885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5917189443119621,
+      "learning_rate": 6.223795440118015e-06,
+      "loss": 0.5172,
+      "step": 6886
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6571296009952919,
+      "learning_rate": 6.2227952046683945e-06,
+      "loss": 0.5163,
+      "step": 6887
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8671543582304182,
+      "learning_rate": 6.22179491716829e-06,
+      "loss": 0.637,
+      "step": 6888
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.599192086806491,
+      "learning_rate": 6.22079457766028e-06,
+      "loss": 0.5182,
+      "step": 6889
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7168495563588103,
+      "learning_rate": 6.219794186186945e-06,
+      "loss": 0.6026,
+      "step": 6890
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6995747763004064,
+      "learning_rate": 6.21879374279087e-06,
+      "loss": 0.5996,
+      "step": 6891
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6169746328907965,
+      "learning_rate": 6.217793247514638e-06,
+      "loss": 0.513,
+      "step": 6892
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2195289350244276,
+      "learning_rate": 6.216792700400841e-06,
+      "loss": 0.6221,
+      "step": 6893
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.749395846257989,
+      "learning_rate": 6.215792101492068e-06,
+      "loss": 0.6589,
+      "step": 6894
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6770192456971336,
+      "learning_rate": 6.214791450830908e-06,
+      "loss": 0.5235,
+      "step": 6895
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5865526980333283,
+      "learning_rate": 6.2137907484599605e-06,
+      "loss": 0.5403,
+      "step": 6896
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8193177300012351,
+      "learning_rate": 6.212789994421817e-06,
+      "loss": 0.6289,
+      "step": 6897
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9135725602561945,
+      "learning_rate": 6.21178918875908e-06,
+      "loss": 0.5734,
+      "step": 6898
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0049178283070277,
+      "learning_rate": 6.210788331514349e-06,
+      "loss": 0.6387,
+      "step": 6899
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8394995515526554,
+      "learning_rate": 6.209787422730229e-06,
+      "loss": 0.6422,
+      "step": 6900
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7368126956076453,
+      "learning_rate": 6.208786462449323e-06,
+      "loss": 0.6024,
+      "step": 6901
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8384003347264457,
+      "learning_rate": 6.207785450714242e-06,
+      "loss": 0.6018,
+      "step": 6902
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7074314150809423,
+      "learning_rate": 6.206784387567592e-06,
+      "loss": 0.5318,
+      "step": 6903
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5627536859957167,
+      "learning_rate": 6.205783273051988e-06,
+      "loss": 0.5226,
+      "step": 6904
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6223252024920127,
+      "learning_rate": 6.204782107210044e-06,
+      "loss": 0.5519,
+      "step": 6905
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6013096013428314,
+      "learning_rate": 6.203780890084374e-06,
+      "loss": 0.4471,
+      "step": 6906
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6376797279349884,
+      "learning_rate": 6.2027796217175985e-06,
+      "loss": 0.5521,
+      "step": 6907
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9504777812926263,
+      "learning_rate": 6.20177830215234e-06,
+      "loss": 0.6301,
+      "step": 6908
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.957302861659713,
+      "learning_rate": 6.200776931431216e-06,
+      "loss": 0.6086,
+      "step": 6909
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9441578880772518,
+      "learning_rate": 6.199775509596857e-06,
+      "loss": 0.655,
+      "step": 6910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7527162940865,
+      "learning_rate": 6.198774036691888e-06,
+      "loss": 0.5941,
+      "step": 6911
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5929544365399939,
+      "learning_rate": 6.197772512758939e-06,
+      "loss": 0.512,
+      "step": 6912
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5774386893457697,
+      "learning_rate": 6.1967709378406425e-06,
+      "loss": 0.5229,
+      "step": 6913
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7774892969808319,
+      "learning_rate": 6.195769311979631e-06,
+      "loss": 0.6513,
+      "step": 6914
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8135022353073249,
+      "learning_rate": 6.194767635218541e-06,
+      "loss": 0.6416,
+      "step": 6915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8896958118115137,
+      "learning_rate": 6.193765907600011e-06,
+      "loss": 0.5814,
+      "step": 6916
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.628869665013241,
+      "learning_rate": 6.192764129166681e-06,
+      "loss": 0.5466,
+      "step": 6917
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5756931612956352,
+      "learning_rate": 6.191762299961194e-06,
+      "loss": 0.5114,
+      "step": 6918
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5808018003374665,
+      "learning_rate": 6.190760420026193e-06,
+      "loss": 0.524,
+      "step": 6919
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6915554516000589,
+      "learning_rate": 6.189758489404327e-06,
+      "loss": 0.5687,
+      "step": 6920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.636161627714552,
+      "learning_rate": 6.1887565081382435e-06,
+      "loss": 0.6299,
+      "step": 6921
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6774435912803495,
+      "learning_rate": 6.187754476270595e-06,
+      "loss": 0.5514,
+      "step": 6922
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8193920191532753,
+      "learning_rate": 6.186752393844032e-06,
+      "loss": 0.5995,
+      "step": 6923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0017311733553744,
+      "learning_rate": 6.185750260901213e-06,
+      "loss": 0.5734,
+      "step": 6924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9726053865363808,
+      "learning_rate": 6.184748077484796e-06,
+      "loss": 0.6573,
+      "step": 6925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7189541834067743,
+      "learning_rate": 6.183745843637437e-06,
+      "loss": 0.594,
+      "step": 6926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7879463764308462,
+      "learning_rate": 6.182743559401801e-06,
+      "loss": 0.664,
+      "step": 6927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6473155588635416,
+      "learning_rate": 6.181741224820552e-06,
+      "loss": 0.4793,
+      "step": 6928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7817179856107083,
+      "learning_rate": 6.180738839936354e-06,
+      "loss": 0.5041,
+      "step": 6929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5812415751280942,
+      "learning_rate": 6.179736404791877e-06,
+      "loss": 0.5268,
+      "step": 6930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6793739797508815,
+      "learning_rate": 6.178733919429793e-06,
+      "loss": 0.5722,
+      "step": 6931
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6494114722260511,
+      "learning_rate": 6.177731383892771e-06,
+      "loss": 0.5401,
+      "step": 6932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5785418031395532,
+      "learning_rate": 6.176728798223488e-06,
+      "loss": 0.5074,
+      "step": 6933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5861779976293513,
+      "learning_rate": 6.1757261624646194e-06,
+      "loss": 0.5502,
+      "step": 6934
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6706083698895646,
+      "learning_rate": 6.174723476658845e-06,
+      "loss": 0.5165,
+      "step": 6935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6137462044331675,
+      "learning_rate": 6.1737207408488474e-06,
+      "loss": 0.539,
+      "step": 6936
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6274706029119311,
+      "learning_rate": 6.172717955077309e-06,
+      "loss": 0.5049,
+      "step": 6937
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6239881438203184,
+      "learning_rate": 6.171715119386913e-06,
+      "loss": 0.5485,
+      "step": 6938
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8374711885935975,
+      "learning_rate": 6.1707122338203496e-06,
+      "loss": 0.576,
+      "step": 6939
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.778003576136719,
+      "learning_rate": 6.169709298420308e-06,
+      "loss": 0.6208,
+      "step": 6940
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6405126665539604,
+      "learning_rate": 6.1687063132294776e-06,
+      "loss": 0.5289,
+      "step": 6941
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7634993817468099,
+      "learning_rate": 6.167703278290556e-06,
+      "loss": 0.6004,
+      "step": 6942
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9057512115608788,
+      "learning_rate": 6.166700193646235e-06,
+      "loss": 0.653,
+      "step": 6943
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5547006559169239,
+      "learning_rate": 6.165697059339218e-06,
+      "loss": 0.5182,
+      "step": 6944
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5080745747688111,
+      "learning_rate": 6.1646938754122e-06,
+      "loss": 0.455,
+      "step": 6945
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5615632718437535,
+      "learning_rate": 6.1636906419078865e-06,
+      "loss": 0.4883,
+      "step": 6946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6593833915263615,
+      "learning_rate": 6.162687358868979e-06,
+      "loss": 0.5638,
+      "step": 6947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8758431849895096,
+      "learning_rate": 6.161684026338188e-06,
+      "loss": 0.5644,
+      "step": 6948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7714982101403007,
+      "learning_rate": 6.160680644358221e-06,
+      "loss": 0.6517,
+      "step": 6949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7229073069281465,
+      "learning_rate": 6.159677212971788e-06,
+      "loss": 0.5827,
+      "step": 6950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7170060430273959,
+      "learning_rate": 6.158673732221601e-06,
+      "loss": 0.6252,
+      "step": 6951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6044232803295256,
+      "learning_rate": 6.157670202150374e-06,
+      "loss": 0.5583,
+      "step": 6952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6175731398380181,
+      "learning_rate": 6.156666622800829e-06,
+      "loss": 0.4854,
+      "step": 6953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.726168022342418,
+      "learning_rate": 6.155662994215679e-06,
+      "loss": 0.6317,
+      "step": 6954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1010979581015248,
+      "learning_rate": 6.15465931643765e-06,
+      "loss": 0.7149,
+      "step": 6955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5951092840843372,
+      "learning_rate": 6.153655589509464e-06,
+      "loss": 0.4564,
+      "step": 6956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7322830194642305,
+      "learning_rate": 6.1526518134738455e-06,
+      "loss": 0.6375,
+      "step": 6957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7118970209441388,
+      "learning_rate": 6.1516479883735234e-06,
+      "loss": 0.5813,
+      "step": 6958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6104637232344424,
+      "learning_rate": 6.150644114251226e-06,
+      "loss": 0.5411,
+      "step": 6959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.613150555109829,
+      "learning_rate": 6.149640191149684e-06,
+      "loss": 0.5812,
+      "step": 6960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5887407967297529,
+      "learning_rate": 6.148636219111635e-06,
+      "loss": 0.5466,
+      "step": 6961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6929918455206895,
+      "learning_rate": 6.147632198179813e-06,
+      "loss": 0.5655,
+      "step": 6962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6886215737073035,
+      "learning_rate": 6.1466281283969545e-06,
+      "loss": 0.5481,
+      "step": 6963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6728231630783966,
+      "learning_rate": 6.1456240098058005e-06,
+      "loss": 0.5506,
+      "step": 6964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.72738416349824,
+      "learning_rate": 6.144619842449094e-06,
+      "loss": 0.5739,
+      "step": 6965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6921865149606771,
+      "learning_rate": 6.143615626369578e-06,
+      "loss": 0.5789,
+      "step": 6966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7517713041007731,
+      "learning_rate": 6.1426113616099995e-06,
+      "loss": 0.603,
+      "step": 6967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7732631404374873,
+      "learning_rate": 6.141607048213107e-06,
+      "loss": 0.6654,
+      "step": 6968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8065041814625559,
+      "learning_rate": 6.14060268622165e-06,
+      "loss": 0.4992,
+      "step": 6969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7775832453012291,
+      "learning_rate": 6.139598275678381e-06,
+      "loss": 0.6143,
+      "step": 6970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8101687830242854,
+      "learning_rate": 6.138593816626055e-06,
+      "loss": 0.5874,
+      "step": 6971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7719784037915997,
+      "learning_rate": 6.137589309107429e-06,
+      "loss": 0.5557,
+      "step": 6972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6989662084373551,
+      "learning_rate": 6.136584753165262e-06,
+      "loss": 0.5359,
+      "step": 6973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.663647152962222,
+      "learning_rate": 6.135580148842314e-06,
+      "loss": 0.5134,
+      "step": 6974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7800516951878073,
+      "learning_rate": 6.134575496181348e-06,
+      "loss": 0.6692,
+      "step": 6975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6679711901363279,
+      "learning_rate": 6.133570795225128e-06,
+      "loss": 0.4585,
+      "step": 6976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6182815652001041,
+      "learning_rate": 6.132566046016422e-06,
+      "loss": 0.5263,
+      "step": 6977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5417296473212366,
+      "learning_rate": 6.131561248597997e-06,
+      "loss": 0.5297,
+      "step": 6978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6054942574934512,
+      "learning_rate": 6.1305564030126264e-06,
+      "loss": 0.5422,
+      "step": 6979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6145197552505882,
+      "learning_rate": 6.1295515093030835e-06,
+      "loss": 0.5241,
+      "step": 6980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6276337869361472,
+      "learning_rate": 6.1285465675121414e-06,
+      "loss": 0.541,
+      "step": 6981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6780973649138582,
+      "learning_rate": 6.1275415776825795e-06,
+      "loss": 0.5792,
+      "step": 6982
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6101682433886345,
+      "learning_rate": 6.1265365398571755e-06,
+      "loss": 0.5689,
+      "step": 6983
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7210033167807183,
+      "learning_rate": 6.1255314540787105e-06,
+      "loss": 0.583,
+      "step": 6984
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7737271743546085,
+      "learning_rate": 6.1245263203899665e-06,
+      "loss": 0.6431,
+      "step": 6985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7812190187635045,
+      "learning_rate": 6.123521138833732e-06,
+      "loss": 0.5542,
+      "step": 6986
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5628385126075388,
+      "learning_rate": 6.122515909452793e-06,
+      "loss": 0.536,
+      "step": 6987
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6866876473192836,
+      "learning_rate": 6.121510632289939e-06,
+      "loss": 0.5375,
+      "step": 6988
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8312761759667844,
+      "learning_rate": 6.1205053073879605e-06,
+      "loss": 0.6062,
+      "step": 6989
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5482930227360178,
+      "learning_rate": 6.119499934789654e-06,
+      "loss": 0.4728,
+      "step": 6990
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6022037195005886,
+      "learning_rate": 6.118494514537809e-06,
+      "loss": 0.5124,
+      "step": 6991
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6006512478769933,
+      "learning_rate": 6.117489046675229e-06,
+      "loss": 0.5193,
+      "step": 6992
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6906369730060452,
+      "learning_rate": 6.11648353124471e-06,
+      "loss": 0.5501,
+      "step": 6993
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7418386740062759,
+      "learning_rate": 6.115477968289057e-06,
+      "loss": 0.553,
+      "step": 6994
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6265433495134833,
+      "learning_rate": 6.114472357851069e-06,
+      "loss": 0.53,
+      "step": 6995
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6038562518554125,
+      "learning_rate": 6.1134666999735555e-06,
+      "loss": 0.5465,
+      "step": 6996
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7887925635533009,
+      "learning_rate": 6.1124609946993215e-06,
+      "loss": 0.6539,
+      "step": 6997
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7531312921636764,
+      "learning_rate": 6.1114552420711795e-06,
+      "loss": 0.5588,
+      "step": 6998
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.786681441606181,
+      "learning_rate": 6.1104494421319385e-06,
+      "loss": 0.6205,
+      "step": 6999
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6395401569832504,
+      "learning_rate": 6.1094435949244135e-06,
+      "loss": 0.5425,
+      "step": 7000
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6192397084103954,
+      "learning_rate": 6.1084377004914195e-06,
+      "loss": 0.5775,
+      "step": 7001
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6416149970841349,
+      "learning_rate": 6.107431758875776e-06,
+      "loss": 0.5446,
+      "step": 7002
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6091182591298585,
+      "learning_rate": 6.106425770120299e-06,
+      "loss": 0.5454,
+      "step": 7003
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6439199776065763,
+      "learning_rate": 6.105419734267815e-06,
+      "loss": 0.514,
+      "step": 7004
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0701801205887538,
+      "learning_rate": 6.104413651361144e-06,
+      "loss": 0.5988,
+      "step": 7005
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.832089162578337,
+      "learning_rate": 6.103407521443115e-06,
+      "loss": 0.5974,
+      "step": 7006
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8201899620076895,
+      "learning_rate": 6.1024013445565515e-06,
+      "loss": 0.6743,
+      "step": 7007
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7319994524681279,
+      "learning_rate": 6.101395120744286e-06,
+      "loss": 0.5104,
+      "step": 7008
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5550222837796425,
+      "learning_rate": 6.100388850049149e-06,
+      "loss": 0.5644,
+      "step": 7009
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5438093201350016,
+      "learning_rate": 6.099382532513976e-06,
+      "loss": 0.5014,
+      "step": 7010
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.715979102804903,
+      "learning_rate": 6.098376168181602e-06,
+      "loss": 0.6365,
+      "step": 7011
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6219142752938432,
+      "learning_rate": 6.097369757094864e-06,
+      "loss": 0.5332,
+      "step": 7012
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5800098704411566,
+      "learning_rate": 6.0963632992966036e-06,
+      "loss": 0.5211,
+      "step": 7013
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6475692842033535,
+      "learning_rate": 6.09535679482966e-06,
+      "loss": 0.5382,
+      "step": 7014
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6349157139075965,
+      "learning_rate": 6.094350243736878e-06,
+      "loss": 0.4815,
+      "step": 7015
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9500280625936495,
+      "learning_rate": 6.093343646061101e-06,
+      "loss": 0.6404,
+      "step": 7016
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5721372743526819,
+      "learning_rate": 6.092337001845181e-06,
+      "loss": 0.5161,
+      "step": 7017
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9799732367864387,
+      "learning_rate": 6.091330311131965e-06,
+      "loss": 0.5629,
+      "step": 7018
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5914921112517635,
+      "learning_rate": 6.090323573964305e-06,
+      "loss": 0.5108,
+      "step": 7019
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5518055965519464,
+      "learning_rate": 6.089316790385056e-06,
+      "loss": 0.4329,
+      "step": 7020
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8094592049857515,
+      "learning_rate": 6.088309960437071e-06,
+      "loss": 0.6236,
+      "step": 7021
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7519590278852772,
+      "learning_rate": 6.087303084163207e-06,
+      "loss": 0.5772,
+      "step": 7022
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6435099144700268,
+      "learning_rate": 6.086296161606328e-06,
+      "loss": 0.504,
+      "step": 7023
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7090464668951559,
+      "learning_rate": 6.085289192809291e-06,
+      "loss": 0.6317,
+      "step": 7024
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5962875285809118,
+      "learning_rate": 6.084282177814962e-06,
+      "loss": 0.5077,
+      "step": 7025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.722109036630549,
+      "learning_rate": 6.083275116666206e-06,
+      "loss": 0.5609,
+      "step": 7026
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7958360727980037,
+      "learning_rate": 6.0822680094058885e-06,
+      "loss": 0.5995,
+      "step": 7027
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5849184702604707,
+      "learning_rate": 6.081260856076882e-06,
+      "loss": 0.4885,
+      "step": 7028
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5601944460241765,
+      "learning_rate": 6.0802536567220546e-06,
+      "loss": 0.4801,
+      "step": 7029
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8724346766325471,
+      "learning_rate": 6.079246411384282e-06,
+      "loss": 0.624,
+      "step": 7030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6521322783295203,
+      "learning_rate": 6.0782391201064374e-06,
+      "loss": 0.4872,
+      "step": 7031
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7911287449441682,
+      "learning_rate": 6.0772317829313994e-06,
+      "loss": 0.6163,
+      "step": 7032
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9396882391719757,
+      "learning_rate": 6.0762243999020465e-06,
+      "loss": 0.5203,
+      "step": 7033
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5714283129024395,
+      "learning_rate": 6.0752169710612605e-06,
+      "loss": 0.5076,
+      "step": 7034
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7593724104306929,
+      "learning_rate": 6.074209496451924e-06,
+      "loss": 0.6125,
+      "step": 7035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6438712476384332,
+      "learning_rate": 6.073201976116923e-06,
+      "loss": 0.5102,
+      "step": 7036
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5416491414306279,
+      "learning_rate": 6.072194410099142e-06,
+      "loss": 0.4788,
+      "step": 7037
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7547145890173839,
+      "learning_rate": 6.07118679844147e-06,
+      "loss": 0.5941,
+      "step": 7038
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6843022383752474,
+      "learning_rate": 6.070179141186802e-06,
+      "loss": 0.5519,
+      "step": 7039
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6489889671240688,
+      "learning_rate": 6.069171438378025e-06,
+      "loss": 0.5307,
+      "step": 7040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6018158642552033,
+      "learning_rate": 6.068163690058038e-06,
+      "loss": 0.4985,
+      "step": 7041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6810515386416672,
+      "learning_rate": 6.067155896269735e-06,
+      "loss": 0.5313,
+      "step": 7042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6795412536103733,
+      "learning_rate": 6.066148057056017e-06,
+      "loss": 0.5429,
+      "step": 7043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5638607829429544,
+      "learning_rate": 6.065140172459782e-06,
+      "loss": 0.4877,
+      "step": 7044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6394401456542274,
+      "learning_rate": 6.064132242523935e-06,
+      "loss": 0.5457,
+      "step": 7045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5914186049789284,
+      "learning_rate": 6.063124267291378e-06,
+      "loss": 0.525,
+      "step": 7046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.794023482227983,
+      "learning_rate": 6.0621162468050165e-06,
+      "loss": 0.5801,
+      "step": 7047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.632927484052232,
+      "learning_rate": 6.061108181107762e-06,
+      "loss": 0.5272,
+      "step": 7048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8562898361911961,
+      "learning_rate": 6.060100070242524e-06,
+      "loss": 0.6606,
+      "step": 7049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6690535977493315,
+      "learning_rate": 6.059091914252213e-06,
+      "loss": 0.5403,
+      "step": 7050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.519046681801265,
+      "learning_rate": 6.058083713179743e-06,
+      "loss": 0.4995,
+      "step": 7051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6134678210055258,
+      "learning_rate": 6.057075467068031e-06,
+      "loss": 0.5696,
+      "step": 7052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7754917113503742,
+      "learning_rate": 6.056067175959993e-06,
+      "loss": 0.5921,
+      "step": 7053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.527923556579244,
+      "learning_rate": 6.055058839898551e-06,
+      "loss": 0.4658,
+      "step": 7054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7248729441933248,
+      "learning_rate": 6.054050458926626e-06,
+      "loss": 0.5888,
+      "step": 7055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8296417343177384,
+      "learning_rate": 6.053042033087141e-06,
+      "loss": 0.6083,
+      "step": 7056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6675126820356825,
+      "learning_rate": 6.052033562423022e-06,
+      "loss": 0.56,
+      "step": 7057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.668743435739018,
+      "learning_rate": 6.051025046977196e-06,
+      "loss": 0.5379,
+      "step": 7058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7949294816777245,
+      "learning_rate": 6.050016486792591e-06,
+      "loss": 0.6224,
+      "step": 7059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6205793233071754,
+      "learning_rate": 6.049007881912141e-06,
+      "loss": 0.5338,
+      "step": 7060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6408703407888654,
+      "learning_rate": 6.047999232378777e-06,
+      "loss": 0.5325,
+      "step": 7061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.729530647059725,
+      "learning_rate": 6.046990538235435e-06,
+      "loss": 0.5263,
+      "step": 7062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8706779640023593,
+      "learning_rate": 6.045981799525051e-06,
+      "loss": 0.6181,
+      "step": 7063
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5438266929594107,
+      "learning_rate": 6.0449730162905654e-06,
+      "loss": 0.4777,
+      "step": 7064
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7359860876038248,
+      "learning_rate": 6.043964188574915e-06,
+      "loss": 0.5942,
+      "step": 7065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7241323835662788,
+      "learning_rate": 6.042955316421049e-06,
+      "loss": 0.6209,
+      "step": 7066
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8035127333265772,
+      "learning_rate": 6.041946399871905e-06,
+      "loss": 0.5655,
+      "step": 7067
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6956269386666362,
+      "learning_rate": 6.0409374389704335e-06,
+      "loss": 0.5964,
+      "step": 7068
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6660078947711036,
+      "learning_rate": 6.039928433759582e-06,
+      "loss": 0.5378,
+      "step": 7069
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.589503771019717,
+      "learning_rate": 6.0389193842823e-06,
+      "loss": 0.5409,
+      "step": 7070
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5866593047121362,
+      "learning_rate": 6.037910290581538e-06,
+      "loss": 0.4917,
+      "step": 7071
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6604860005466706,
+      "learning_rate": 6.036901152700253e-06,
+      "loss": 0.5242,
+      "step": 7072
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5987571440606564,
+      "learning_rate": 6.0358919706814e-06,
+      "loss": 0.5085,
+      "step": 7073
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6245365173020011,
+      "learning_rate": 6.034882744567936e-06,
+      "loss": 0.4873,
+      "step": 7074
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6262500080672581,
+      "learning_rate": 6.033873474402819e-06,
+      "loss": 0.4865,
+      "step": 7075
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6391414222372909,
+      "learning_rate": 6.032864160229014e-06,
+      "loss": 0.5861,
+      "step": 7076
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6128183206857968,
+      "learning_rate": 6.0318548020894805e-06,
+      "loss": 0.5292,
+      "step": 7077
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5810153720405562,
+      "learning_rate": 6.030845400027186e-06,
+      "loss": 0.553,
+      "step": 7078
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7575050037271156,
+      "learning_rate": 6.029835954085097e-06,
+      "loss": 0.6126,
+      "step": 7079
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7596357862222697,
+      "learning_rate": 6.028826464306183e-06,
+      "loss": 0.6089,
+      "step": 7080
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5339339948938192,
+      "learning_rate": 6.027816930733413e-06,
+      "loss": 0.4795,
+      "step": 7081
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8948840595780623,
+      "learning_rate": 6.026807353409762e-06,
+      "loss": 0.6023,
+      "step": 7082
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.732670892906231,
+      "learning_rate": 6.0257977323782025e-06,
+      "loss": 0.6055,
+      "step": 7083
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6642119407639585,
+      "learning_rate": 6.02478806768171e-06,
+      "loss": 0.5293,
+      "step": 7084
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5666997521276964,
+      "learning_rate": 6.023778359363266e-06,
+      "loss": 0.4599,
+      "step": 7085
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8067941820503703,
+      "learning_rate": 6.022768607465849e-06,
+      "loss": 0.6445,
+      "step": 7086
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8362104587225783,
+      "learning_rate": 6.02175881203244e-06,
+      "loss": 0.6197,
+      "step": 7087
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9487357113369056,
+      "learning_rate": 6.0207489731060234e-06,
+      "loss": 0.6509,
+      "step": 7088
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6394645143540107,
+      "learning_rate": 6.019739090729585e-06,
+      "loss": 0.5156,
+      "step": 7089
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7150408067804788,
+      "learning_rate": 6.018729164946112e-06,
+      "loss": 0.5964,
+      "step": 7090
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8315369821989891,
+      "learning_rate": 6.017719195798595e-06,
+      "loss": 0.6406,
+      "step": 7091
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.699505475163175,
+      "learning_rate": 6.016709183330023e-06,
+      "loss": 0.5979,
+      "step": 7092
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6470538483120143,
+      "learning_rate": 6.0156991275833895e-06,
+      "loss": 0.5236,
+      "step": 7093
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.816930821266561,
+      "learning_rate": 6.0146890286016905e-06,
+      "loss": 0.5391,
+      "step": 7094
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7199362070824941,
+      "learning_rate": 6.013678886427921e-06,
+      "loss": 0.5074,
+      "step": 7095
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7627784379619833,
+      "learning_rate": 6.012668701105081e-06,
+      "loss": 0.531,
+      "step": 7096
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0517068494106159,
+      "learning_rate": 6.011658472676172e-06,
+      "loss": 0.6331,
+      "step": 7097
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7076510314830814,
+      "learning_rate": 6.010648201184193e-06,
+      "loss": 0.6351,
+      "step": 7098
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7192164568048685,
+      "learning_rate": 6.009637886672151e-06,
+      "loss": 0.6225,
+      "step": 7099
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6591453296503997,
+      "learning_rate": 6.008627529183049e-06,
+      "loss": 0.5122,
+      "step": 7100
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.908526565693428,
+      "learning_rate": 6.007617128759897e-06,
+      "loss": 0.636,
+      "step": 7101
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6526237649533455,
+      "learning_rate": 6.006606685445703e-06,
+      "loss": 0.5644,
+      "step": 7102
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5511515448112384,
+      "learning_rate": 6.005596199283479e-06,
+      "loss": 0.4747,
+      "step": 7103
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6527267519955694,
+      "learning_rate": 6.004585670316239e-06,
+      "loss": 0.5206,
+      "step": 7104
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7818144627686768,
+      "learning_rate": 6.003575098586997e-06,
+      "loss": 0.6209,
+      "step": 7105
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6316059920579868,
+      "learning_rate": 6.00256448413877e-06,
+      "loss": 0.5712,
+      "step": 7106
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5420774402223564,
+      "learning_rate": 6.001553827014577e-06,
+      "loss": 0.4941,
+      "step": 7107
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6586518366308904,
+      "learning_rate": 6.000543127257438e-06,
+      "loss": 0.5324,
+      "step": 7108
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7115921284145044,
+      "learning_rate": 5.999532384910374e-06,
+      "loss": 0.5458,
+      "step": 7109
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7526653039661527,
+      "learning_rate": 5.998521600016411e-06,
+      "loss": 0.5702,
+      "step": 7110
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8628833770275489,
+      "learning_rate": 5.997510772618576e-06,
+      "loss": 0.6212,
+      "step": 7111
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7805227832120558,
+      "learning_rate": 5.9964999027598935e-06,
+      "loss": 0.5973,
+      "step": 7112
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5602003628481105,
+      "learning_rate": 5.995488990483395e-06,
+      "loss": 0.5268,
+      "step": 7113
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6177402675696717,
+      "learning_rate": 5.994478035832111e-06,
+      "loss": 0.5986,
+      "step": 7114
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5909271036335947,
+      "learning_rate": 5.993467038849075e-06,
+      "loss": 0.5722,
+      "step": 7115
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8137966181961045,
+      "learning_rate": 5.9924559995773215e-06,
+      "loss": 0.6335,
+      "step": 7116
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8055655690688112,
+      "learning_rate": 5.991444918059887e-06,
+      "loss": 0.6479,
+      "step": 7117
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7368075426446393,
+      "learning_rate": 5.990433794339812e-06,
+      "loss": 0.5828,
+      "step": 7118
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6376893108332394,
+      "learning_rate": 5.9894226284601355e-06,
+      "loss": 0.5768,
+      "step": 7119
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7633560778863631,
+      "learning_rate": 5.988411420463898e-06,
+      "loss": 0.6285,
+      "step": 7120
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5656994053803676,
+      "learning_rate": 5.987400170394146e-06,
+      "loss": 0.5029,
+      "step": 7121
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6560430090900162,
+      "learning_rate": 5.986388878293923e-06,
+      "loss": 0.5222,
+      "step": 7122
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7305105623059527,
+      "learning_rate": 5.985377544206278e-06,
+      "loss": 0.5946,
+      "step": 7123
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7100536238867052,
+      "learning_rate": 5.98436616817426e-06,
+      "loss": 0.5535,
+      "step": 7124
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3688451111270026,
+      "learning_rate": 5.98335475024092e-06,
+      "loss": 0.5401,
+      "step": 7125
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.606096566139521,
+      "learning_rate": 5.982343290449311e-06,
+      "loss": 0.5654,
+      "step": 7126
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9467145391972902,
+      "learning_rate": 5.981331788842485e-06,
+      "loss": 0.5792,
+      "step": 7127
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8465251048556427,
+      "learning_rate": 5.980320245463502e-06,
+      "loss": 0.6222,
+      "step": 7128
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5689033131252232,
+      "learning_rate": 5.979308660355419e-06,
+      "loss": 0.4894,
+      "step": 7129
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5929282360825587,
+      "learning_rate": 5.978297033561295e-06,
+      "loss": 0.5315,
+      "step": 7130
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2609677645202362,
+      "learning_rate": 5.977285365124195e-06,
+      "loss": 0.6,
+      "step": 7131
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6110118892147859,
+      "learning_rate": 5.976273655087178e-06,
+      "loss": 0.4886,
+      "step": 7132
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7791127968859274,
+      "learning_rate": 5.97526190349331e-06,
+      "loss": 0.619,
+      "step": 7133
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5583853862987631,
+      "learning_rate": 5.974250110385661e-06,
+      "loss": 0.5145,
+      "step": 7134
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6382480622421793,
+      "learning_rate": 5.9732382758072985e-06,
+      "loss": 0.5536,
+      "step": 7135
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.824847088257256,
+      "learning_rate": 5.9722263998012905e-06,
+      "loss": 0.6172,
+      "step": 7136
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7809217260638395,
+      "learning_rate": 5.971214482410713e-06,
+      "loss": 0.619,
+      "step": 7137
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9115552827998088,
+      "learning_rate": 5.9702025236786385e-06,
+      "loss": 0.6926,
+      "step": 7138
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8096508120858464,
+      "learning_rate": 5.969190523648143e-06,
+      "loss": 0.5813,
+      "step": 7139
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5385794462804069,
+      "learning_rate": 5.9681784823623035e-06,
+      "loss": 0.4845,
+      "step": 7140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5701950137838097,
+      "learning_rate": 5.967166399864199e-06,
+      "loss": 0.5091,
+      "step": 7141
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6510968743689732,
+      "learning_rate": 5.9661542761969134e-06,
+      "loss": 0.5597,
+      "step": 7142
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6148001912811577,
+      "learning_rate": 5.965142111403527e-06,
+      "loss": 0.5745,
+      "step": 7143
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6375709545722904,
+      "learning_rate": 5.964129905527125e-06,
+      "loss": 0.5938,
+      "step": 7144
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8125509007726149,
+      "learning_rate": 5.963117658610794e-06,
+      "loss": 0.5854,
+      "step": 7145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.616028726265624,
+      "learning_rate": 5.96210537069762e-06,
+      "loss": 0.4918,
+      "step": 7146
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6498585460254964,
+      "learning_rate": 5.961093041830698e-06,
+      "loss": 0.5581,
+      "step": 7147
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.739770037097316,
+      "learning_rate": 5.960080672053115e-06,
+      "loss": 0.5302,
+      "step": 7148
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8914153248838461,
+      "learning_rate": 5.959068261407965e-06,
+      "loss": 0.6369,
+      "step": 7149
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7522871068945128,
+      "learning_rate": 5.958055809938345e-06,
+      "loss": 0.5893,
+      "step": 7150
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8661673027769349,
+      "learning_rate": 5.95704331768735e-06,
+      "loss": 0.6221,
+      "step": 7151
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8137186745809454,
+      "learning_rate": 5.956030784698081e-06,
+      "loss": 0.6102,
+      "step": 7152
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6389780260807015,
+      "learning_rate": 5.9550182110136345e-06,
+      "loss": 0.4756,
+      "step": 7153
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9007838809121117,
+      "learning_rate": 5.954005596677115e-06,
+      "loss": 0.6118,
+      "step": 7154
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8722979637220367,
+      "learning_rate": 5.952992941731626e-06,
+      "loss": 0.6217,
+      "step": 7155
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6290544796256371,
+      "learning_rate": 5.951980246220272e-06,
+      "loss": 0.5266,
+      "step": 7156
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7740339630176876,
+      "learning_rate": 5.9509675101861604e-06,
+      "loss": 0.6114,
+      "step": 7157
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7094801167297725,
+      "learning_rate": 5.9499547336724025e-06,
+      "loss": 0.5376,
+      "step": 7158
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.670911401387463,
+      "learning_rate": 5.948941916722107e-06,
+      "loss": 0.5188,
+      "step": 7159
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6805473534722848,
+      "learning_rate": 5.9479290593783865e-06,
+      "loss": 0.5408,
+      "step": 7160
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5415560859130355,
+      "learning_rate": 5.9469161616843554e-06,
+      "loss": 0.5114,
+      "step": 7161
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6999988889602705,
+      "learning_rate": 5.945903223683128e-06,
+      "loss": 0.5516,
+      "step": 7162
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5969446972468377,
+      "learning_rate": 5.944890245417825e-06,
+      "loss": 0.5858,
+      "step": 7163
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.780556428227834,
+      "learning_rate": 5.94387722693156e-06,
+      "loss": 0.6036,
+      "step": 7164
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5729227086510631,
+      "learning_rate": 5.942864168267461e-06,
+      "loss": 0.5475,
+      "step": 7165
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5526932724807357,
+      "learning_rate": 5.941851069468646e-06,
+      "loss": 0.53,
+      "step": 7166
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6565265821662218,
+      "learning_rate": 5.9408379305782415e-06,
+      "loss": 0.5202,
+      "step": 7167
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6865929034117233,
+      "learning_rate": 5.939824751639373e-06,
+      "loss": 0.5723,
+      "step": 7168
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7925018445090999,
+      "learning_rate": 5.938811532695166e-06,
+      "loss": 0.5967,
+      "step": 7169
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7102343774352992,
+      "learning_rate": 5.937798273788754e-06,
+      "loss": 0.5892,
+      "step": 7170
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8779940712735889,
+      "learning_rate": 5.936784974963266e-06,
+      "loss": 0.5544,
+      "step": 7171
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6772792441690239,
+      "learning_rate": 5.935771636261835e-06,
+      "loss": 0.5276,
+      "step": 7172
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.043407671374952,
+      "learning_rate": 5.934758257727595e-06,
+      "loss": 0.5983,
+      "step": 7173
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5830818131182096,
+      "learning_rate": 5.933744839403683e-06,
+      "loss": 0.4846,
+      "step": 7174
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8823314896482773,
+      "learning_rate": 5.932731381333239e-06,
+      "loss": 0.646,
+      "step": 7175
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6289780052938609,
+      "learning_rate": 5.931717883559399e-06,
+      "loss": 0.52,
+      "step": 7176
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5476421579791547,
+      "learning_rate": 5.930704346125306e-06,
+      "loss": 0.5383,
+      "step": 7177
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.635112023277828,
+      "learning_rate": 5.929690769074103e-06,
+      "loss": 0.5503,
+      "step": 7178
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8178462896208889,
+      "learning_rate": 5.928677152448935e-06,
+      "loss": 0.6054,
+      "step": 7179
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7534776143426254,
+      "learning_rate": 5.9276634962929495e-06,
+      "loss": 0.6388,
+      "step": 7180
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7323572319169261,
+      "learning_rate": 5.926649800649293e-06,
+      "loss": 0.5689,
+      "step": 7181
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9494818224270936,
+      "learning_rate": 5.925636065561113e-06,
+      "loss": 0.5651,
+      "step": 7182
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7334370926046292,
+      "learning_rate": 5.9246222910715655e-06,
+      "loss": 0.5804,
+      "step": 7183
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1804973622481583,
+      "learning_rate": 5.923608477223803e-06,
+      "loss": 0.593,
+      "step": 7184
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5964010756064292,
+      "learning_rate": 5.922594624060978e-06,
+      "loss": 0.5087,
+      "step": 7185
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.753845475829064,
+      "learning_rate": 5.921580731626248e-06,
+      "loss": 0.5902,
+      "step": 7186
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6525589041353362,
+      "learning_rate": 5.92056679996277e-06,
+      "loss": 0.5023,
+      "step": 7187
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6700678120394333,
+      "learning_rate": 5.919552829113707e-06,
+      "loss": 0.5163,
+      "step": 7188
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7170149945197423,
+      "learning_rate": 5.918538819122217e-06,
+      "loss": 0.5363,
+      "step": 7189
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8488018843512644,
+      "learning_rate": 5.917524770031465e-06,
+      "loss": 0.592,
+      "step": 7190
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6533620222904847,
+      "learning_rate": 5.916510681884616e-06,
+      "loss": 0.5723,
+      "step": 7191
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6408270049772504,
+      "learning_rate": 5.915496554724837e-06,
+      "loss": 0.4436,
+      "step": 7192
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6786165644820878,
+      "learning_rate": 5.914482388595294e-06,
+      "loss": 0.5396,
+      "step": 7193
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6308099791159926,
+      "learning_rate": 5.913468183539158e-06,
+      "loss": 0.5553,
+      "step": 7194
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5258517317425928,
+      "learning_rate": 5.9124539395996e-06,
+      "loss": 0.5148,
+      "step": 7195
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6638753437974643,
+      "learning_rate": 5.911439656819794e-06,
+      "loss": 0.5294,
+      "step": 7196
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5745802024832648,
+      "learning_rate": 5.910425335242914e-06,
+      "loss": 0.5142,
+      "step": 7197
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7218634303445264,
+      "learning_rate": 5.9094109749121375e-06,
+      "loss": 0.592,
+      "step": 7198
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8275995525581679,
+      "learning_rate": 5.9083965758706415e-06,
+      "loss": 0.6407,
+      "step": 7199
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7125587368919056,
+      "learning_rate": 5.907382138161607e-06,
+      "loss": 0.6005,
+      "step": 7200
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.683047625340476,
+      "learning_rate": 5.906367661828214e-06,
+      "loss": 0.5889,
+      "step": 7201
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8188590980388989,
+      "learning_rate": 5.905353146913645e-06,
+      "loss": 0.6049,
+      "step": 7202
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5829997877623249,
+      "learning_rate": 5.904338593461087e-06,
+      "loss": 0.544,
+      "step": 7203
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0746563096274913,
+      "learning_rate": 5.903324001513724e-06,
+      "loss": 0.5755,
+      "step": 7204
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6011755085581385,
+      "learning_rate": 5.902309371114745e-06,
+      "loss": 0.5377,
+      "step": 7205
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6765894960483733,
+      "learning_rate": 5.901294702307339e-06,
+      "loss": 0.5617,
+      "step": 7206
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6419749281686844,
+      "learning_rate": 5.900279995134699e-06,
+      "loss": 0.5236,
+      "step": 7207
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7260466237244242,
+      "learning_rate": 5.899265249640014e-06,
+      "loss": 0.5627,
+      "step": 7208
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6804589382568738,
+      "learning_rate": 5.898250465866483e-06,
+      "loss": 0.6041,
+      "step": 7209
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7260330565911145,
+      "learning_rate": 5.8972356438573e-06,
+      "loss": 0.6468,
+      "step": 7210
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8120694239781333,
+      "learning_rate": 5.896220783655663e-06,
+      "loss": 0.6234,
+      "step": 7211
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7537635229403995,
+      "learning_rate": 5.89520588530477e-06,
+      "loss": 0.6105,
+      "step": 7212
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.631604668002357,
+      "learning_rate": 5.894190948847824e-06,
+      "loss": 0.5171,
+      "step": 7213
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.613916991971185,
+      "learning_rate": 5.893175974328027e-06,
+      "loss": 0.5376,
+      "step": 7214
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6614329216719724,
+      "learning_rate": 5.892160961788582e-06,
+      "loss": 0.5349,
+      "step": 7215
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6898093405177034,
+      "learning_rate": 5.891145911272697e-06,
+      "loss": 0.5927,
+      "step": 7216
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2641626101636871,
+      "learning_rate": 5.890130822823578e-06,
+      "loss": 0.6472,
+      "step": 7217
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5193522151683049,
+      "learning_rate": 5.889115696484433e-06,
+      "loss": 0.4806,
+      "step": 7218
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.693461264309939,
+      "learning_rate": 5.888100532298474e-06,
+      "loss": 0.5587,
+      "step": 7219
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.775924265788248,
+      "learning_rate": 5.8870853303089145e-06,
+      "loss": 0.6299,
+      "step": 7220
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7564253579679279,
+      "learning_rate": 5.886070090558967e-06,
+      "loss": 0.6224,
+      "step": 7221
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7511805664848867,
+      "learning_rate": 5.885054813091847e-06,
+      "loss": 0.545,
+      "step": 7222
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7532209394603636,
+      "learning_rate": 5.884039497950773e-06,
+      "loss": 0.6317,
+      "step": 7223
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6535111507141641,
+      "learning_rate": 5.883024145178961e-06,
+      "loss": 0.5001,
+      "step": 7224
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6102915383448848,
+      "learning_rate": 5.882008754819634e-06,
+      "loss": 0.5686,
+      "step": 7225
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8643434331530937,
+      "learning_rate": 5.880993326916012e-06,
+      "loss": 0.6236,
+      "step": 7226
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7603682620802733,
+      "learning_rate": 5.879977861511319e-06,
+      "loss": 0.5728,
+      "step": 7227
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7440511129750577,
+      "learning_rate": 5.878962358648781e-06,
+      "loss": 0.5748,
+      "step": 7228
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7453367244684165,
+      "learning_rate": 5.877946818371624e-06,
+      "loss": 0.593,
+      "step": 7229
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7142201825666366,
+      "learning_rate": 5.876931240723076e-06,
+      "loss": 0.5748,
+      "step": 7230
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6348020574814905,
+      "learning_rate": 5.875915625746369e-06,
+      "loss": 0.4652,
+      "step": 7231
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5732572572783668,
+      "learning_rate": 5.874899973484731e-06,
+      "loss": 0.5243,
+      "step": 7232
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8021992513079979,
+      "learning_rate": 5.8738842839813966e-06,
+      "loss": 0.5826,
+      "step": 7233
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5565199083502078,
+      "learning_rate": 5.8728685572796025e-06,
+      "loss": 0.4745,
+      "step": 7234
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0035041655435828,
+      "learning_rate": 5.871852793422582e-06,
+      "loss": 0.6218,
+      "step": 7235
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.564083561414303,
+      "learning_rate": 5.870836992453576e-06,
+      "loss": 0.5022,
+      "step": 7236
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8312163668841487,
+      "learning_rate": 5.86982115441582e-06,
+      "loss": 0.5982,
+      "step": 7237
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.556786251097832,
+      "learning_rate": 5.868805279352559e-06,
+      "loss": 0.5339,
+      "step": 7238
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7339868029490866,
+      "learning_rate": 5.867789367307031e-06,
+      "loss": 0.6447,
+      "step": 7239
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7161191679438755,
+      "learning_rate": 5.8667734183224835e-06,
+      "loss": 0.5506,
+      "step": 7240
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7817509442462736,
+      "learning_rate": 5.865757432442162e-06,
+      "loss": 0.6232,
+      "step": 7241
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8237965183738162,
+      "learning_rate": 5.864741409709313e-06,
+      "loss": 0.6474,
+      "step": 7242
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7999948748300804,
+      "learning_rate": 5.863725350167185e-06,
+      "loss": 0.6375,
+      "step": 7243
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6992666595332188,
+      "learning_rate": 5.8627092538590305e-06,
+      "loss": 0.6001,
+      "step": 7244
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6423170497339399,
+      "learning_rate": 5.861693120828097e-06,
+      "loss": 0.5058,
+      "step": 7245
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8022773476160895,
+      "learning_rate": 5.860676951117643e-06,
+      "loss": 0.587,
+      "step": 7246
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7104105558828214,
+      "learning_rate": 5.859660744770922e-06,
+      "loss": 0.527,
+      "step": 7247
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7307800551745314,
+      "learning_rate": 5.858644501831189e-06,
+      "loss": 0.5882,
+      "step": 7248
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8297109238648962,
+      "learning_rate": 5.857628222341705e-06,
+      "loss": 0.5899,
+      "step": 7249
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6407306982382849,
+      "learning_rate": 5.856611906345726e-06,
+      "loss": 0.5241,
+      "step": 7250
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5248318932797091,
+      "learning_rate": 5.855595553886516e-06,
+      "loss": 0.4604,
+      "step": 7251
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7399990124653967,
+      "learning_rate": 5.854579165007338e-06,
+      "loss": 0.4952,
+      "step": 7252
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7943497501869831,
+      "learning_rate": 5.853562739751455e-06,
+      "loss": 0.6566,
+      "step": 7253
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6319138225181162,
+      "learning_rate": 5.852546278162135e-06,
+      "loss": 0.5243,
+      "step": 7254
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8630142886579614,
+      "learning_rate": 5.851529780282643e-06,
+      "loss": 0.5679,
+      "step": 7255
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.648310338628721,
+      "learning_rate": 5.850513246156251e-06,
+      "loss": 0.5886,
+      "step": 7256
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5592694301524513,
+      "learning_rate": 5.849496675826226e-06,
+      "loss": 0.5162,
+      "step": 7257
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6810846159214108,
+      "learning_rate": 5.848480069335843e-06,
+      "loss": 0.5328,
+      "step": 7258
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.892395292191634,
+      "learning_rate": 5.847463426728375e-06,
+      "loss": 0.6174,
+      "step": 7259
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.825818034903631,
+      "learning_rate": 5.846446748047098e-06,
+      "loss": 0.6505,
+      "step": 7260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7165267176836811,
+      "learning_rate": 5.845430033335286e-06,
+      "loss": 0.5006,
+      "step": 7261
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6705033128727618,
+      "learning_rate": 5.84441328263622e-06,
+      "loss": 0.5677,
+      "step": 7262
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6894400756286736,
+      "learning_rate": 5.843396495993179e-06,
+      "loss": 0.4957,
+      "step": 7263
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7164755467312727,
+      "learning_rate": 5.842379673449443e-06,
+      "loss": 0.6148,
+      "step": 7264
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6902510171666226,
+      "learning_rate": 5.841362815048297e-06,
+      "loss": 0.5347,
+      "step": 7265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6157792280727161,
+      "learning_rate": 5.840345920833025e-06,
+      "loss": 0.5318,
+      "step": 7266
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6523571112720159,
+      "learning_rate": 5.839328990846913e-06,
+      "loss": 0.5343,
+      "step": 7267
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6491774948464895,
+      "learning_rate": 5.838312025133247e-06,
+      "loss": 0.5205,
+      "step": 7268
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6413892153133484,
+      "learning_rate": 5.837295023735318e-06,
+      "loss": 0.5371,
+      "step": 7269
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5681557721160956,
+      "learning_rate": 5.836277986696413e-06,
+      "loss": 0.5094,
+      "step": 7270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7352275222706279,
+      "learning_rate": 5.835260914059828e-06,
+      "loss": 0.5759,
+      "step": 7271
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6516485592214992,
+      "learning_rate": 5.8342438058688554e-06,
+      "loss": 0.5103,
+      "step": 7272
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8861948647087583,
+      "learning_rate": 5.83322666216679e-06,
+      "loss": 0.5509,
+      "step": 7273
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1264271629169584,
+      "learning_rate": 5.832209482996927e-06,
+      "loss": 0.6502,
+      "step": 7274
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6007806751201401,
+      "learning_rate": 5.8311922684025665e-06,
+      "loss": 0.5171,
+      "step": 7275
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5815624149811355,
+      "learning_rate": 5.830175018427007e-06,
+      "loss": 0.5103,
+      "step": 7276
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.746432587254956,
+      "learning_rate": 5.829157733113551e-06,
+      "loss": 0.6755,
+      "step": 7277
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5971462350187131,
+      "learning_rate": 5.828140412505499e-06,
+      "loss": 0.4912,
+      "step": 7278
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6938170524870781,
+      "learning_rate": 5.827123056646156e-06,
+      "loss": 0.5943,
+      "step": 7279
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6160802896861488,
+      "learning_rate": 5.826105665578827e-06,
+      "loss": 0.4987,
+      "step": 7280
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6498898330890953,
+      "learning_rate": 5.82508823934682e-06,
+      "loss": 0.5708,
+      "step": 7281
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7738949323289743,
+      "learning_rate": 5.8240707779934435e-06,
+      "loss": 0.5498,
+      "step": 7282
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7765301423890085,
+      "learning_rate": 5.823053281562008e-06,
+      "loss": 0.623,
+      "step": 7283
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7132515388069783,
+      "learning_rate": 5.822035750095824e-06,
+      "loss": 0.6325,
+      "step": 7284
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7143228425316213,
+      "learning_rate": 5.821018183638204e-06,
+      "loss": 0.5357,
+      "step": 7285
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5636076813919914,
+      "learning_rate": 5.820000582232465e-06,
+      "loss": 0.4801,
+      "step": 7286
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.794985452295735,
+      "learning_rate": 5.818982945921921e-06,
+      "loss": 0.5896,
+      "step": 7287
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6115182909504777,
+      "learning_rate": 5.8179652747498885e-06,
+      "loss": 0.4978,
+      "step": 7288
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7767621973329313,
+      "learning_rate": 5.8169475687596885e-06,
+      "loss": 0.6628,
+      "step": 7289
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6632843083244028,
+      "learning_rate": 5.8159298279946415e-06,
+      "loss": 0.5125,
+      "step": 7290
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5658385117973802,
+      "learning_rate": 5.814912052498071e-06,
+      "loss": 0.4695,
+      "step": 7291
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5294647688395007,
+      "learning_rate": 5.813894242313297e-06,
+      "loss": 0.5064,
+      "step": 7292
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5538328673703952,
+      "learning_rate": 5.812876397483645e-06,
+      "loss": 0.4584,
+      "step": 7293
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6773857563583625,
+      "learning_rate": 5.811858518052445e-06,
+      "loss": 0.5655,
+      "step": 7294
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8299489916019177,
+      "learning_rate": 5.810840604063019e-06,
+      "loss": 0.5688,
+      "step": 7295
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8512504343072516,
+      "learning_rate": 5.809822655558701e-06,
+      "loss": 0.6078,
+      "step": 7296
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6345573429900008,
+      "learning_rate": 5.808804672582821e-06,
+      "loss": 0.5575,
+      "step": 7297
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5266777328024713,
+      "learning_rate": 5.80778665517871e-06,
+      "loss": 0.4691,
+      "step": 7298
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7904841558684887,
+      "learning_rate": 5.806768603389703e-06,
+      "loss": 0.6036,
+      "step": 7299
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.610487168331742,
+      "learning_rate": 5.805750517259134e-06,
+      "loss": 0.5656,
+      "step": 7300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.777017542457991,
+      "learning_rate": 5.8047323968303395e-06,
+      "loss": 0.571,
+      "step": 7301
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5665595055161324,
+      "learning_rate": 5.803714242146659e-06,
+      "loss": 0.464,
+      "step": 7302
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6278771042895364,
+      "learning_rate": 5.802696053251432e-06,
+      "loss": 0.555,
+      "step": 7303
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6927763676822408,
+      "learning_rate": 5.801677830187999e-06,
+      "loss": 0.6089,
+      "step": 7304
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6001409500907885,
+      "learning_rate": 5.800659572999703e-06,
+      "loss": 0.5146,
+      "step": 7305
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6843889087530002,
+      "learning_rate": 5.799641281729887e-06,
+      "loss": 0.5852,
+      "step": 7306
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6261159346037729,
+      "learning_rate": 5.798622956421897e-06,
+      "loss": 0.5592,
+      "step": 7307
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.710718794672907,
+      "learning_rate": 5.79760459711908e-06,
+      "loss": 0.6004,
+      "step": 7308
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5953656751254692,
+      "learning_rate": 5.796586203864784e-06,
+      "loss": 0.5419,
+      "step": 7309
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5661716398805003,
+      "learning_rate": 5.795567776702358e-06,
+      "loss": 0.5404,
+      "step": 7310
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7715856442709891,
+      "learning_rate": 5.794549315675155e-06,
+      "loss": 0.6305,
+      "step": 7311
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6470501375160036,
+      "learning_rate": 5.793530820826526e-06,
+      "loss": 0.5366,
+      "step": 7312
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6474857512999157,
+      "learning_rate": 5.792512292199825e-06,
+      "loss": 0.5427,
+      "step": 7313
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.558137666083832,
+      "learning_rate": 5.79149372983841e-06,
+      "loss": 0.4815,
+      "step": 7314
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6000185161865772,
+      "learning_rate": 5.790475133785636e-06,
+      "loss": 0.5203,
+      "step": 7315
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8911390778156995,
+      "learning_rate": 5.789456504084861e-06,
+      "loss": 0.543,
+      "step": 7316
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7606267790585807,
+      "learning_rate": 5.788437840779445e-06,
+      "loss": 0.5925,
+      "step": 7317
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7153722157555635,
+      "learning_rate": 5.78741914391275e-06,
+      "loss": 0.6094,
+      "step": 7318
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8110151344299823,
+      "learning_rate": 5.786400413528137e-06,
+      "loss": 0.5681,
+      "step": 7319
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9068505354730736,
+      "learning_rate": 5.785381649668973e-06,
+      "loss": 0.6366,
+      "step": 7320
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7482058628075415,
+      "learning_rate": 5.7843628523786224e-06,
+      "loss": 0.6049,
+      "step": 7321
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7868115218222387,
+      "learning_rate": 5.78334402170045e-06,
+      "loss": 0.7053,
+      "step": 7322
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5478806940317313,
+      "learning_rate": 5.782325157677827e-06,
+      "loss": 0.5363,
+      "step": 7323
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6553736676047508,
+      "learning_rate": 5.781306260354121e-06,
+      "loss": 0.4501,
+      "step": 7324
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6139132501769308,
+      "learning_rate": 5.780287329772705e-06,
+      "loss": 0.5402,
+      "step": 7325
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6355656960870817,
+      "learning_rate": 5.77926836597695e-06,
+      "loss": 0.4826,
+      "step": 7326
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7906631153158185,
+      "learning_rate": 5.778249369010231e-06,
+      "loss": 0.6307,
+      "step": 7327
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6654108367236921,
+      "learning_rate": 5.777230338915925e-06,
+      "loss": 0.5398,
+      "step": 7328
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7802883108848935,
+      "learning_rate": 5.776211275737404e-06,
+      "loss": 0.6146,
+      "step": 7329
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8730210217566167,
+      "learning_rate": 5.775192179518052e-06,
+      "loss": 0.6861,
+      "step": 7330
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6642755256253636,
+      "learning_rate": 5.774173050301246e-06,
+      "loss": 0.5598,
+      "step": 7331
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6828931354082131,
+      "learning_rate": 5.773153888130365e-06,
+      "loss": 0.5477,
+      "step": 7332
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8504980540566135,
+      "learning_rate": 5.772134693048796e-06,
+      "loss": 0.6737,
+      "step": 7333
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8626703537939912,
+      "learning_rate": 5.771115465099919e-06,
+      "loss": 0.6658,
+      "step": 7334
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7064141432567075,
+      "learning_rate": 5.770096204327121e-06,
+      "loss": 0.5683,
+      "step": 7335
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7370490035885552,
+      "learning_rate": 5.76907691077379e-06,
+      "loss": 0.5936,
+      "step": 7336
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7928807843531035,
+      "learning_rate": 5.768057584483311e-06,
+      "loss": 0.6099,
+      "step": 7337
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5556098517357636,
+      "learning_rate": 5.767038225499075e-06,
+      "loss": 0.4458,
+      "step": 7338
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5269786739704576,
+      "learning_rate": 5.766018833864474e-06,
+      "loss": 0.4881,
+      "step": 7339
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6518402603283743,
+      "learning_rate": 5.764999409622899e-06,
+      "loss": 0.593,
+      "step": 7340
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8607622627960236,
+      "learning_rate": 5.763979952817742e-06,
+      "loss": 0.6132,
+      "step": 7341
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5550745442235996,
+      "learning_rate": 5.762960463492402e-06,
+      "loss": 0.4642,
+      "step": 7342
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6982318183972722,
+      "learning_rate": 5.761940941690271e-06,
+      "loss": 0.5902,
+      "step": 7343
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6883622646587318,
+      "learning_rate": 5.76092138745475e-06,
+      "loss": 0.517,
+      "step": 7344
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6213778884618992,
+      "learning_rate": 5.759901800829236e-06,
+      "loss": 0.5233,
+      "step": 7345
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6308561300454176,
+      "learning_rate": 5.758882181857132e-06,
+      "loss": 0.5012,
+      "step": 7346
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6492288607943988,
+      "learning_rate": 5.757862530581838e-06,
+      "loss": 0.5539,
+      "step": 7347
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5186728290297209,
+      "learning_rate": 5.7568428470467585e-06,
+      "loss": 0.5327,
+      "step": 7348
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6570017285976666,
+      "learning_rate": 5.755823131295297e-06,
+      "loss": 0.5791,
+      "step": 7349
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8145197848347757,
+      "learning_rate": 5.7548033833708594e-06,
+      "loss": 0.6382,
+      "step": 7350
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6820084692761272,
+      "learning_rate": 5.753783603316854e-06,
+      "loss": 0.5583,
+      "step": 7351
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6332458317116423,
+      "learning_rate": 5.75276379117669e-06,
+      "loss": 0.5421,
+      "step": 7352
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6461586618587962,
+      "learning_rate": 5.7517439469937775e-06,
+      "loss": 0.5205,
+      "step": 7353
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6096986912375691,
+      "learning_rate": 5.750724070811526e-06,
+      "loss": 0.5177,
+      "step": 7354
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7751609533970542,
+      "learning_rate": 5.74970416267335e-06,
+      "loss": 0.6546,
+      "step": 7355
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5793601989944073,
+      "learning_rate": 5.748684222622664e-06,
+      "loss": 0.476,
+      "step": 7356
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.636761912942386,
+      "learning_rate": 5.747664250702882e-06,
+      "loss": 0.5679,
+      "step": 7357
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5912910977846981,
+      "learning_rate": 5.746644246957423e-06,
+      "loss": 0.4883,
+      "step": 7358
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8115293024724972,
+      "learning_rate": 5.745624211429705e-06,
+      "loss": 0.5448,
+      "step": 7359
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6331336748521982,
+      "learning_rate": 5.744604144163146e-06,
+      "loss": 0.5274,
+      "step": 7360
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6858756021816557,
+      "learning_rate": 5.7435840452011695e-06,
+      "loss": 0.548,
+      "step": 7361
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5814533625575896,
+      "learning_rate": 5.742563914587195e-06,
+      "loss": 0.4806,
+      "step": 7362
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8448686031097746,
+      "learning_rate": 5.741543752364646e-06,
+      "loss": 0.604,
+      "step": 7363
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7131375390963478,
+      "learning_rate": 5.740523558576951e-06,
+      "loss": 0.6207,
+      "step": 7364
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6041193615922887,
+      "learning_rate": 5.739503333267535e-06,
+      "loss": 0.5577,
+      "step": 7365
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6476494627256795,
+      "learning_rate": 5.738483076479825e-06,
+      "loss": 0.5687,
+      "step": 7366
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5732051931699436,
+      "learning_rate": 5.73746278825725e-06,
+      "loss": 0.5384,
+      "step": 7367
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7014535737222668,
+      "learning_rate": 5.73644246864324e-06,
+      "loss": 0.5415,
+      "step": 7368
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.890259532896133,
+      "learning_rate": 5.735422117681228e-06,
+      "loss": 0.6049,
+      "step": 7369
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.962262608558948,
+      "learning_rate": 5.734401735414646e-06,
+      "loss": 0.6252,
+      "step": 7370
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6922243929075977,
+      "learning_rate": 5.733381321886929e-06,
+      "loss": 0.6154,
+      "step": 7371
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8310460094501119,
+      "learning_rate": 5.732360877141514e-06,
+      "loss": 0.5878,
+      "step": 7372
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6793760578554305,
+      "learning_rate": 5.731340401221835e-06,
+      "loss": 0.595,
+      "step": 7373
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6752946060588411,
+      "learning_rate": 5.730319894171335e-06,
+      "loss": 0.5282,
+      "step": 7374
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6121113321309787,
+      "learning_rate": 5.729299356033446e-06,
+      "loss": 0.5057,
+      "step": 7375
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6289692836393576,
+      "learning_rate": 5.728278786851618e-06,
+      "loss": 0.5292,
+      "step": 7376
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5610746626640373,
+      "learning_rate": 5.727258186669288e-06,
+      "loss": 0.4688,
+      "step": 7377
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.627680869265941,
+      "learning_rate": 5.726237555529901e-06,
+      "loss": 0.5053,
+      "step": 7378
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6137069671782773,
+      "learning_rate": 5.7252168934769024e-06,
+      "loss": 0.501,
+      "step": 7379
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7840440910327866,
+      "learning_rate": 5.724196200553738e-06,
+      "loss": 0.5845,
+      "step": 7380
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7814997338324008,
+      "learning_rate": 5.723175476803854e-06,
+      "loss": 0.6358,
+      "step": 7381
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8715585755067938,
+      "learning_rate": 5.722154722270703e-06,
+      "loss": 0.6125,
+      "step": 7382
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6487162391170347,
+      "learning_rate": 5.721133936997732e-06,
+      "loss": 0.5712,
+      "step": 7383
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.541181427380261,
+      "learning_rate": 5.720113121028394e-06,
+      "loss": 0.4945,
+      "step": 7384
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.645279531852808,
+      "learning_rate": 5.719092274406142e-06,
+      "loss": 0.5564,
+      "step": 7385
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8028404859534641,
+      "learning_rate": 5.718071397174429e-06,
+      "loss": 0.606,
+      "step": 7386
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6000211867106248,
+      "learning_rate": 5.717050489376712e-06,
+      "loss": 0.4833,
+      "step": 7387
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6481284086082985,
+      "learning_rate": 5.7160295510564456e-06,
+      "loss": 0.4924,
+      "step": 7388
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6870893518460691,
+      "learning_rate": 5.715008582257091e-06,
+      "loss": 0.5259,
+      "step": 7389
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5624763214807417,
+      "learning_rate": 5.713987583022106e-06,
+      "loss": 0.4652,
+      "step": 7390
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7734190247354849,
+      "learning_rate": 5.71296655339495e-06,
+      "loss": 0.5363,
+      "step": 7391
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6086705751030543,
+      "learning_rate": 5.7119454934190866e-06,
+      "loss": 0.564,
+      "step": 7392
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6276722535599929,
+      "learning_rate": 5.710924403137979e-06,
+      "loss": 0.5594,
+      "step": 7393
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7243127724305656,
+      "learning_rate": 5.70990328259509e-06,
+      "loss": 0.6064,
+      "step": 7394
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6605308434412986,
+      "learning_rate": 5.708882131833888e-06,
+      "loss": 0.5117,
+      "step": 7395
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0070714308613156,
+      "learning_rate": 5.707860950897839e-06,
+      "loss": 0.531,
+      "step": 7396
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7500848913152038,
+      "learning_rate": 5.706839739830411e-06,
+      "loss": 0.5478,
+      "step": 7397
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5618937124625564,
+      "learning_rate": 5.705818498675074e-06,
+      "loss": 0.4884,
+      "step": 7398
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6870789565856263,
+      "learning_rate": 5.704797227475299e-06,
+      "loss": 0.4901,
+      "step": 7399
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5456379918592325,
+      "learning_rate": 5.703775926274559e-06,
+      "loss": 0.5161,
+      "step": 7400
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7181685879281244,
+      "learning_rate": 5.702754595116325e-06,
+      "loss": 0.6051,
+      "step": 7401
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6961210746654559,
+      "learning_rate": 5.701733234044075e-06,
+      "loss": 0.5463,
+      "step": 7402
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6056251106359543,
+      "learning_rate": 5.700711843101283e-06,
+      "loss": 0.5379,
+      "step": 7403
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5872243201901712,
+      "learning_rate": 5.699690422331426e-06,
+      "loss": 0.5413,
+      "step": 7404
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5540002123578797,
+      "learning_rate": 5.698668971777985e-06,
+      "loss": 0.503,
+      "step": 7405
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8072134370319362,
+      "learning_rate": 5.697647491484439e-06,
+      "loss": 0.6224,
+      "step": 7406
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5572745526912098,
+      "learning_rate": 5.696625981494268e-06,
+      "loss": 0.4906,
+      "step": 7407
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6504237339968888,
+      "learning_rate": 5.695604441850955e-06,
+      "loss": 0.5197,
+      "step": 7408
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5274329768422829,
+      "learning_rate": 5.694582872597984e-06,
+      "loss": 0.4954,
+      "step": 7409
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.928078603726346,
+      "learning_rate": 5.69356127377884e-06,
+      "loss": 0.6473,
+      "step": 7410
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9812196709788339,
+      "learning_rate": 5.692539645437009e-06,
+      "loss": 0.6126,
+      "step": 7411
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7582274940934607,
+      "learning_rate": 5.691517987615976e-06,
+      "loss": 0.6435,
+      "step": 7412
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5944706589682655,
+      "learning_rate": 5.690496300359234e-06,
+      "loss": 0.5122,
+      "step": 7413
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6105810228456057,
+      "learning_rate": 5.68947458371027e-06,
+      "loss": 0.5468,
+      "step": 7414
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.65134934033479,
+      "learning_rate": 5.688452837712577e-06,
+      "loss": 0.4929,
+      "step": 7415
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9131277089851544,
+      "learning_rate": 5.687431062409647e-06,
+      "loss": 0.5851,
+      "step": 7416
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8079190453130916,
+      "learning_rate": 5.686409257844973e-06,
+      "loss": 0.5972,
+      "step": 7417
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8850021987073383,
+      "learning_rate": 5.685387424062051e-06,
+      "loss": 0.601,
+      "step": 7418
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7905063975038278,
+      "learning_rate": 5.684365561104375e-06,
+      "loss": 0.6438,
+      "step": 7419
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8213264804998532,
+      "learning_rate": 5.683343669015444e-06,
+      "loss": 0.5828,
+      "step": 7420
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6870240698263178,
+      "learning_rate": 5.682321747838758e-06,
+      "loss": 0.5946,
+      "step": 7421
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6722639333025503,
+      "learning_rate": 5.681299797617815e-06,
+      "loss": 0.5828,
+      "step": 7422
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6818675552627741,
+      "learning_rate": 5.680277818396117e-06,
+      "loss": 0.5748,
+      "step": 7423
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6275135559303036,
+      "learning_rate": 5.679255810217167e-06,
+      "loss": 0.4726,
+      "step": 7424
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8163185683563445,
+      "learning_rate": 5.678233773124465e-06,
+      "loss": 0.6493,
+      "step": 7425
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7144144806646856,
+      "learning_rate": 5.6772117071615206e-06,
+      "loss": 0.5964,
+      "step": 7426
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7516934859178865,
+      "learning_rate": 5.676189612371837e-06,
+      "loss": 0.6067,
+      "step": 7427
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5715928842235714,
+      "learning_rate": 5.675167488798924e-06,
+      "loss": 0.5164,
+      "step": 7428
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8984798482547672,
+      "learning_rate": 5.674145336486287e-06,
+      "loss": 0.6277,
+      "step": 7429
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9353574284200804,
+      "learning_rate": 5.673123155477438e-06,
+      "loss": 0.6345,
+      "step": 7430
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6148041296681221,
+      "learning_rate": 5.672100945815887e-06,
+      "loss": 0.5372,
+      "step": 7431
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9274117995181222,
+      "learning_rate": 5.671078707545147e-06,
+      "loss": 0.6141,
+      "step": 7432
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2818584754791995,
+      "learning_rate": 5.67005644070873e-06,
+      "loss": 0.5977,
+      "step": 7433
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5917544842531124,
+      "learning_rate": 5.6690341453501515e-06,
+      "loss": 0.5296,
+      "step": 7434
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6038382946577495,
+      "learning_rate": 5.668011821512929e-06,
+      "loss": 0.4771,
+      "step": 7435
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5992428069467104,
+      "learning_rate": 5.666989469240576e-06,
+      "loss": 0.5612,
+      "step": 7436
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5447866721389806,
+      "learning_rate": 5.665967088576613e-06,
+      "loss": 0.4725,
+      "step": 7437
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7557642963332849,
+      "learning_rate": 5.664944679564559e-06,
+      "loss": 0.5562,
+      "step": 7438
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9271305379046939,
+      "learning_rate": 5.663922242247936e-06,
+      "loss": 0.6213,
+      "step": 7439
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5811107774774571,
+      "learning_rate": 5.6628997766702644e-06,
+      "loss": 0.512,
+      "step": 7440
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6768434141643868,
+      "learning_rate": 5.6618772828750675e-06,
+      "loss": 0.5404,
+      "step": 7441
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5977941622466924,
+      "learning_rate": 5.660854760905869e-06,
+      "loss": 0.5048,
+      "step": 7442
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.763565937469353,
+      "learning_rate": 5.659832210806195e-06,
+      "loss": 0.6167,
+      "step": 7443
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8373060044331139,
+      "learning_rate": 5.6588096326195726e-06,
+      "loss": 0.6458,
+      "step": 7444
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6689128337283314,
+      "learning_rate": 5.6577870263895306e-06,
+      "loss": 0.4912,
+      "step": 7445
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7595481738334218,
+      "learning_rate": 5.6567643921595965e-06,
+      "loss": 0.6323,
+      "step": 7446
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5712844219189369,
+      "learning_rate": 5.655741729973301e-06,
+      "loss": 0.542,
+      "step": 7447
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5786996986577434,
+      "learning_rate": 5.654719039874175e-06,
+      "loss": 0.4855,
+      "step": 7448
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7366752618305401,
+      "learning_rate": 5.653696321905752e-06,
+      "loss": 0.6589,
+      "step": 7449
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5312668299918883,
+      "learning_rate": 5.6526735761115635e-06,
+      "loss": 0.5062,
+      "step": 7450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8934322478513086,
+      "learning_rate": 5.651650802535149e-06,
+      "loss": 0.6418,
+      "step": 7451
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6447074771875847,
+      "learning_rate": 5.650628001220041e-06,
+      "loss": 0.5331,
+      "step": 7452
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7473002167035475,
+      "learning_rate": 5.6496051722097785e-06,
+      "loss": 0.5632,
+      "step": 7453
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.723231306865618,
+      "learning_rate": 5.648582315547901e-06,
+      "loss": 0.5889,
+      "step": 7454
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7775924834948316,
+      "learning_rate": 5.647559431277944e-06,
+      "loss": 0.5167,
+      "step": 7455
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7407817777605376,
+      "learning_rate": 5.646536519443453e-06,
+      "loss": 0.5996,
+      "step": 7456
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6986294728385855,
+      "learning_rate": 5.645513580087968e-06,
+      "loss": 0.6367,
+      "step": 7457
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.597406972525196,
+      "learning_rate": 5.644490613255034e-06,
+      "loss": 0.5207,
+      "step": 7458
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.711716046917632,
+      "learning_rate": 5.643467618988192e-06,
+      "loss": 0.627,
+      "step": 7459
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8041144330129631,
+      "learning_rate": 5.642444597330992e-06,
+      "loss": 0.5939,
+      "step": 7460
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7501084350677372,
+      "learning_rate": 5.6414215483269764e-06,
+      "loss": 0.5278,
+      "step": 7461
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6171344466764241,
+      "learning_rate": 5.640398472019697e-06,
+      "loss": 0.5095,
+      "step": 7462
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6353046548901394,
+      "learning_rate": 5.6393753684526995e-06,
+      "loss": 0.5005,
+      "step": 7463
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7233307148465343,
+      "learning_rate": 5.638352237669537e-06,
+      "loss": 0.5639,
+      "step": 7464
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6222856119738878,
+      "learning_rate": 5.637329079713758e-06,
+      "loss": 0.4991,
+      "step": 7465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5494478442453682,
+      "learning_rate": 5.636305894628917e-06,
+      "loss": 0.4983,
+      "step": 7466
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9094628837538682,
+      "learning_rate": 5.635282682458568e-06,
+      "loss": 0.6231,
+      "step": 7467
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.571839861934092,
+      "learning_rate": 5.634259443246266e-06,
+      "loss": 0.5163,
+      "step": 7468
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7829450278475736,
+      "learning_rate": 5.633236177035566e-06,
+      "loss": 0.6332,
+      "step": 7469
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9182791455086887,
+      "learning_rate": 5.632212883870024e-06,
+      "loss": 0.6257,
+      "step": 7470
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6367248681491957,
+      "learning_rate": 5.631189563793201e-06,
+      "loss": 0.5602,
+      "step": 7471
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.592248131703522,
+      "learning_rate": 5.630166216848656e-06,
+      "loss": 0.5561,
+      "step": 7472
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6855158527421498,
+      "learning_rate": 5.629142843079948e-06,
+      "loss": 0.6157,
+      "step": 7473
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6711081524201122,
+      "learning_rate": 5.6281194425306386e-06,
+      "loss": 0.5747,
+      "step": 7474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9488468902943121,
+      "learning_rate": 5.627096015244292e-06,
+      "loss": 0.5873,
+      "step": 7475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5924103778816938,
+      "learning_rate": 5.626072561264473e-06,
+      "loss": 0.5409,
+      "step": 7476
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6976615426673333,
+      "learning_rate": 5.625049080634746e-06,
+      "loss": 0.5064,
+      "step": 7477
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7410077206632754,
+      "learning_rate": 5.624025573398676e-06,
+      "loss": 0.564,
+      "step": 7478
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7758628136137053,
+      "learning_rate": 5.623002039599832e-06,
+      "loss": 0.5449,
+      "step": 7479
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.56500355210866,
+      "learning_rate": 5.6219784792817825e-06,
+      "loss": 0.536,
+      "step": 7480
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8655068984818446,
+      "learning_rate": 5.6209548924880954e-06,
+      "loss": 0.6632,
+      "step": 7481
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6461287796399382,
+      "learning_rate": 5.619931279262343e-06,
+      "loss": 0.5459,
+      "step": 7482
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7727957998907783,
+      "learning_rate": 5.618907639648098e-06,
+      "loss": 0.6538,
+      "step": 7483
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8576024726657329,
+      "learning_rate": 5.617883973688931e-06,
+      "loss": 0.6779,
+      "step": 7484
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5699404090759821,
+      "learning_rate": 5.61686028142842e-06,
+      "loss": 0.5295,
+      "step": 7485
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6045262952351791,
+      "learning_rate": 5.615836562910136e-06,
+      "loss": 0.4957,
+      "step": 7486
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6796026296394505,
+      "learning_rate": 5.614812818177657e-06,
+      "loss": 0.558,
+      "step": 7487
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7804945500216487,
+      "learning_rate": 5.6137890472745625e-06,
+      "loss": 0.6621,
+      "step": 7488
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9389280388329666,
+      "learning_rate": 5.61276525024443e-06,
+      "loss": 0.6651,
+      "step": 7489
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.723619080503985,
+      "learning_rate": 5.611741427130838e-06,
+      "loss": 0.5922,
+      "step": 7490
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4457476941116572,
+      "learning_rate": 5.610717577977369e-06,
+      "loss": 0.6387,
+      "step": 7491
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7561310635183401,
+      "learning_rate": 5.609693702827605e-06,
+      "loss": 0.5484,
+      "step": 7492
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6382758271178987,
+      "learning_rate": 5.608669801725125e-06,
+      "loss": 0.4749,
+      "step": 7493
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.659774268820042,
+      "learning_rate": 5.607645874713519e-06,
+      "loss": 0.5292,
+      "step": 7494
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6203594822082383,
+      "learning_rate": 5.60662192183637e-06,
+      "loss": 0.5286,
+      "step": 7495
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6710073324384463,
+      "learning_rate": 5.605597943137264e-06,
+      "loss": 0.5618,
+      "step": 7496
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5949885700201625,
+      "learning_rate": 5.6045739386597885e-06,
+      "loss": 0.5116,
+      "step": 7497
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6406614047485684,
+      "learning_rate": 5.603549908447533e-06,
+      "loss": 0.5583,
+      "step": 7498
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5361097158972228,
+      "learning_rate": 5.602525852544085e-06,
+      "loss": 0.4929,
+      "step": 7499
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7498696685528496,
+      "learning_rate": 5.6015017709930385e-06,
+      "loss": 0.6039,
+      "step": 7500
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8061247815788423,
+      "learning_rate": 5.600477663837983e-06,
+      "loss": 0.5824,
+      "step": 7501
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5361879481965269,
+      "learning_rate": 5.599453531122513e-06,
+      "loss": 0.5044,
+      "step": 7502
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6590883410830884,
+      "learning_rate": 5.5984293728902205e-06,
+      "loss": 0.5061,
+      "step": 7503
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6569733490302738,
+      "learning_rate": 5.597405189184702e-06,
+      "loss": 0.5126,
+      "step": 7504
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5854814873595723,
+      "learning_rate": 5.5963809800495535e-06,
+      "loss": 0.55,
+      "step": 7505
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6750277894763069,
+      "learning_rate": 5.595356745528374e-06,
+      "loss": 0.5024,
+      "step": 7506
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5940732070440541,
+      "learning_rate": 5.59433248566476e-06,
+      "loss": 0.5455,
+      "step": 7507
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6332071317074104,
+      "learning_rate": 5.593308200502311e-06,
+      "loss": 0.5383,
+      "step": 7508
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.79302800554713,
+      "learning_rate": 5.5922838900846275e-06,
+      "loss": 0.5952,
+      "step": 7509
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.757567081335012,
+      "learning_rate": 5.591259554455311e-06,
+      "loss": 0.6282,
+      "step": 7510
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8590191699846675,
+      "learning_rate": 5.590235193657965e-06,
+      "loss": 0.6224,
+      "step": 7511
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7424039259481556,
+      "learning_rate": 5.5892108077361925e-06,
+      "loss": 0.5549,
+      "step": 7512
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6279115641151283,
+      "learning_rate": 5.5881863967336e-06,
+      "loss": 0.547,
+      "step": 7513
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7931095274600247,
+      "learning_rate": 5.5871619606937925e-06,
+      "loss": 0.625,
+      "step": 7514
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6995865900119276,
+      "learning_rate": 5.586137499660376e-06,
+      "loss": 0.6061,
+      "step": 7515
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6616400129795555,
+      "learning_rate": 5.585113013676958e-06,
+      "loss": 0.586,
+      "step": 7516
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8003756150682517,
+      "learning_rate": 5.584088502787151e-06,
+      "loss": 0.5592,
+      "step": 7517
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7594131286486338,
+      "learning_rate": 5.58306396703456e-06,
+      "loss": 0.5948,
+      "step": 7518
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.701980554897668,
+      "learning_rate": 5.5820394064628e-06,
+      "loss": 0.609,
+      "step": 7519
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6053234960436575,
+      "learning_rate": 5.5810148211154835e-06,
+      "loss": 0.4991,
+      "step": 7520
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8351394011794444,
+      "learning_rate": 5.579990211036223e-06,
+      "loss": 0.6658,
+      "step": 7521
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6514037457401015,
+      "learning_rate": 5.578965576268632e-06,
+      "loss": 0.5554,
+      "step": 7522
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6118776643235391,
+      "learning_rate": 5.577940916856326e-06,
+      "loss": 0.504,
+      "step": 7523
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6012276938343325,
+      "learning_rate": 5.576916232842923e-06,
+      "loss": 0.4909,
+      "step": 7524
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6231388550961715,
+      "learning_rate": 5.5758915242720395e-06,
+      "loss": 0.5479,
+      "step": 7525
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7024457676788421,
+      "learning_rate": 5.5748667911872935e-06,
+      "loss": 0.5909,
+      "step": 7526
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6972453079687018,
+      "learning_rate": 5.573842033632305e-06,
+      "loss": 0.6119,
+      "step": 7527
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7642496430911038,
+      "learning_rate": 5.572817251650695e-06,
+      "loss": 0.6192,
+      "step": 7528
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8392541463793172,
+      "learning_rate": 5.571792445286085e-06,
+      "loss": 0.6536,
+      "step": 7529
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5518529616944446,
+      "learning_rate": 5.570767614582098e-06,
+      "loss": 0.5169,
+      "step": 7530
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8068960512228883,
+      "learning_rate": 5.569742759582357e-06,
+      "loss": 0.6492,
+      "step": 7531
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8396541307872485,
+      "learning_rate": 5.568717880330489e-06,
+      "loss": 0.6041,
+      "step": 7532
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6797223660958093,
+      "learning_rate": 5.567692976870117e-06,
+      "loss": 0.584,
+      "step": 7533
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.698770788248921,
+      "learning_rate": 5.56666804924487e-06,
+      "loss": 0.5687,
+      "step": 7534
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7028067449758266,
+      "learning_rate": 5.5656430974983735e-06,
+      "loss": 0.6244,
+      "step": 7535
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8473567542161662,
+      "learning_rate": 5.564618121674258e-06,
+      "loss": 0.5934,
+      "step": 7536
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5850820347804717,
+      "learning_rate": 5.563593121816155e-06,
+      "loss": 0.5086,
+      "step": 7537
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6427982414322996,
+      "learning_rate": 5.562568097967691e-06,
+      "loss": 0.6103,
+      "step": 7538
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6724865274200578,
+      "learning_rate": 5.5615430501725035e-06,
+      "loss": 0.5032,
+      "step": 7539
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.652842206270668,
+      "learning_rate": 5.560517978474222e-06,
+      "loss": 0.4933,
+      "step": 7540
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6119030785598977,
+      "learning_rate": 5.5594928829164804e-06,
+      "loss": 0.501,
+      "step": 7541
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8612807450827021,
+      "learning_rate": 5.558467763542916e-06,
+      "loss": 0.5677,
+      "step": 7542
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5175732781417085,
+      "learning_rate": 5.557442620397162e-06,
+      "loss": 0.4908,
+      "step": 7543
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7201772136975116,
+      "learning_rate": 5.556417453522858e-06,
+      "loss": 0.5323,
+      "step": 7544
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6234394421718971,
+      "learning_rate": 5.55539226296364e-06,
+      "loss": 0.5189,
+      "step": 7545
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.624635148078209,
+      "learning_rate": 5.554367048763149e-06,
+      "loss": 0.5214,
+      "step": 7546
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5900965532139819,
+      "learning_rate": 5.553341810965025e-06,
+      "loss": 0.5515,
+      "step": 7547
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7721391585343826,
+      "learning_rate": 5.552316549612907e-06,
+      "loss": 0.6673,
+      "step": 7548
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7244636016195506,
+      "learning_rate": 5.551291264750439e-06,
+      "loss": 0.5335,
+      "step": 7549
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.764917421470679,
+      "learning_rate": 5.550265956421263e-06,
+      "loss": 0.6519,
+      "step": 7550
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6327877244351915,
+      "learning_rate": 5.549240624669025e-06,
+      "loss": 0.5064,
+      "step": 7551
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6637650168820303,
+      "learning_rate": 5.548215269537368e-06,
+      "loss": 0.5701,
+      "step": 7552
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6466215346331949,
+      "learning_rate": 5.5471898910699385e-06,
+      "loss": 0.5233,
+      "step": 7553
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7797419064637255,
+      "learning_rate": 5.546164489310384e-06,
+      "loss": 0.6099,
+      "step": 7554
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6021766400263207,
+      "learning_rate": 5.545139064302352e-06,
+      "loss": 0.5427,
+      "step": 7555
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.640741001885347,
+      "learning_rate": 5.544113616089493e-06,
+      "loss": 0.5065,
+      "step": 7556
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2158678722939376,
+      "learning_rate": 5.543088144715455e-06,
+      "loss": 0.6242,
+      "step": 7557
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9070171850791727,
+      "learning_rate": 5.542062650223892e-06,
+      "loss": 0.6381,
+      "step": 7558
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6833671135940843,
+      "learning_rate": 5.5410371326584525e-06,
+      "loss": 0.5495,
+      "step": 7559
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7459351868089814,
+      "learning_rate": 5.540011592062791e-06,
+      "loss": 0.5814,
+      "step": 7560
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8477969766915834,
+      "learning_rate": 5.538986028480563e-06,
+      "loss": 0.6157,
+      "step": 7561
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.664010709265732,
+      "learning_rate": 5.537960441955421e-06,
+      "loss": 0.5162,
+      "step": 7562
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.603346770470461,
+      "learning_rate": 5.536934832531022e-06,
+      "loss": 0.5331,
+      "step": 7563
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7180540870225373,
+      "learning_rate": 5.535909200251024e-06,
+      "loss": 0.6083,
+      "step": 7564
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5268191736598359,
+      "learning_rate": 5.534883545159085e-06,
+      "loss": 0.4378,
+      "step": 7565
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.648713835068757,
+      "learning_rate": 5.5338578672988606e-06,
+      "loss": 0.5791,
+      "step": 7566
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5553914496705349,
+      "learning_rate": 5.532832166714011e-06,
+      "loss": 0.519,
+      "step": 7567
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6280371686105825,
+      "learning_rate": 5.531806443448202e-06,
+      "loss": 0.5551,
+      "step": 7568
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6024276775023125,
+      "learning_rate": 5.530780697545091e-06,
+      "loss": 0.5464,
+      "step": 7569
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6414987537819092,
+      "learning_rate": 5.529754929048342e-06,
+      "loss": 0.5025,
+      "step": 7570
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6149237158979458,
+      "learning_rate": 5.528729138001619e-06,
+      "loss": 0.4912,
+      "step": 7571
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5677021856388555,
+      "learning_rate": 5.5277033244485855e-06,
+      "loss": 0.4997,
+      "step": 7572
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7354021725784786,
+      "learning_rate": 5.526677488432909e-06,
+      "loss": 0.611,
+      "step": 7573
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6210531429893793,
+      "learning_rate": 5.525651629998252e-06,
+      "loss": 0.522,
+      "step": 7574
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.585476354979317,
+      "learning_rate": 5.5246257491882875e-06,
+      "loss": 0.5378,
+      "step": 7575
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6296899660493586,
+      "learning_rate": 5.5235998460466796e-06,
+      "loss": 0.5058,
+      "step": 7576
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.591274840977478,
+      "learning_rate": 5.522573920617101e-06,
+      "loss": 0.5376,
+      "step": 7577
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5831344897862408,
+      "learning_rate": 5.52154797294322e-06,
+      "loss": 0.5258,
+      "step": 7578
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0086384210930563,
+      "learning_rate": 5.520522003068709e-06,
+      "loss": 0.5897,
+      "step": 7579
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8012128775170618,
+      "learning_rate": 5.5194960110372376e-06,
+      "loss": 0.6236,
+      "step": 7580
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6149023152075509,
+      "learning_rate": 5.518469996892483e-06,
+      "loss": 0.576,
+      "step": 7581
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6130289171762597,
+      "learning_rate": 5.517443960678116e-06,
+      "loss": 0.5589,
+      "step": 7582
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5702572281951223,
+      "learning_rate": 5.516417902437814e-06,
+      "loss": 0.5385,
+      "step": 7583
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6507560903497862,
+      "learning_rate": 5.515391822215252e-06,
+      "loss": 0.4782,
+      "step": 7584
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7516982480369816,
+      "learning_rate": 5.514365720054108e-06,
+      "loss": 0.5578,
+      "step": 7585
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5204808405538552,
+      "learning_rate": 5.513339595998058e-06,
+      "loss": 0.4889,
+      "step": 7586
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6956055398196305,
+      "learning_rate": 5.5123134500907825e-06,
+      "loss": 0.5202,
+      "step": 7587
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8328162300187069,
+      "learning_rate": 5.51128728237596e-06,
+      "loss": 0.5557,
+      "step": 7588
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9534275043609638,
+      "learning_rate": 5.5102610928972725e-06,
+      "loss": 0.555,
+      "step": 7589
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.579336801661149,
+      "learning_rate": 5.5092348816983985e-06,
+      "loss": 0.5375,
+      "step": 7590
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9391063810120132,
+      "learning_rate": 5.508208648823025e-06,
+      "loss": 0.5938,
+      "step": 7591
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9695965226750517,
+      "learning_rate": 5.507182394314832e-06,
+      "loss": 0.6178,
+      "step": 7592
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9005124611626333,
+      "learning_rate": 5.506156118217506e-06,
+      "loss": 0.6008,
+      "step": 7593
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.48799709277974174,
+      "learning_rate": 5.5051298205747315e-06,
+      "loss": 0.4413,
+      "step": 7594
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7910818204108412,
+      "learning_rate": 5.504103501430195e-06,
+      "loss": 0.5167,
+      "step": 7595
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5549774493898005,
+      "learning_rate": 5.503077160827583e-06,
+      "loss": 0.5302,
+      "step": 7596
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8722196089232908,
+      "learning_rate": 5.502050798810584e-06,
+      "loss": 0.6307,
+      "step": 7597
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7201682416213975,
+      "learning_rate": 5.501024415422885e-06,
+      "loss": 0.6432,
+      "step": 7598
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8569120142915779,
+      "learning_rate": 5.499998010708181e-06,
+      "loss": 0.5749,
+      "step": 7599
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7118726448812661,
+      "learning_rate": 5.498971584710158e-06,
+      "loss": 0.5539,
+      "step": 7600
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.596970360923443,
+      "learning_rate": 5.497945137472508e-06,
+      "loss": 0.5225,
+      "step": 7601
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8129042638483244,
+      "learning_rate": 5.496918669038927e-06,
+      "loss": 0.6055,
+      "step": 7602
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8967171840855497,
+      "learning_rate": 5.495892179453104e-06,
+      "loss": 0.5558,
+      "step": 7603
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8305055714431683,
+      "learning_rate": 5.494865668758737e-06,
+      "loss": 0.5358,
+      "step": 7604
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6550216912099499,
+      "learning_rate": 5.493839136999517e-06,
+      "loss": 0.5422,
+      "step": 7605
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5836326198558194,
+      "learning_rate": 5.492812584219146e-06,
+      "loss": 0.4814,
+      "step": 7606
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6999942862132048,
+      "learning_rate": 5.491786010461317e-06,
+      "loss": 0.5101,
+      "step": 7607
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5963761150255326,
+      "learning_rate": 5.49075941576973e-06,
+      "loss": 0.5577,
+      "step": 7608
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5399962748439829,
+      "learning_rate": 5.4897328001880825e-06,
+      "loss": 0.494,
+      "step": 7609
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8377025018325649,
+      "learning_rate": 5.488706163760075e-06,
+      "loss": 0.6595,
+      "step": 7610
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6355270949141507,
+      "learning_rate": 5.487679506529405e-06,
+      "loss": 0.5234,
+      "step": 7611
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8021011466859977,
+      "learning_rate": 5.48665282853978e-06,
+      "loss": 0.6364,
+      "step": 7612
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6488854554999794,
+      "learning_rate": 5.485626129834898e-06,
+      "loss": 0.5075,
+      "step": 7613
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5743535463440743,
+      "learning_rate": 5.484599410458464e-06,
+      "loss": 0.5091,
+      "step": 7614
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5576089792025481,
+      "learning_rate": 5.483572670454181e-06,
+      "loss": 0.4852,
+      "step": 7615
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5250051752808631,
+      "learning_rate": 5.482545909865755e-06,
+      "loss": 0.533,
+      "step": 7616
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6262217878642798,
+      "learning_rate": 5.4815191287368914e-06,
+      "loss": 0.5604,
+      "step": 7617
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5474155704323684,
+      "learning_rate": 5.480492327111298e-06,
+      "loss": 0.5137,
+      "step": 7618
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.603400583681847,
+      "learning_rate": 5.4794655050326806e-06,
+      "loss": 0.5085,
+      "step": 7619
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7122199404116781,
+      "learning_rate": 5.4784386625447495e-06,
+      "loss": 0.5079,
+      "step": 7620
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5957349064292433,
+      "learning_rate": 5.477411799691213e-06,
+      "loss": 0.5013,
+      "step": 7621
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5806603744617725,
+      "learning_rate": 5.476384916515783e-06,
+      "loss": 0.5218,
+      "step": 7622
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7985190895949581,
+      "learning_rate": 5.475358013062167e-06,
+      "loss": 0.6171,
+      "step": 7623
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8152396987787083,
+      "learning_rate": 5.474331089374081e-06,
+      "loss": 0.606,
+      "step": 7624
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7558770416267557,
+      "learning_rate": 5.473304145495236e-06,
+      "loss": 0.5792,
+      "step": 7625
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6950817632840202,
+      "learning_rate": 5.472277181469346e-06,
+      "loss": 0.5081,
+      "step": 7626
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7726956635313686,
+      "learning_rate": 5.471250197340127e-06,
+      "loss": 0.6511,
+      "step": 7627
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5827437905910543,
+      "learning_rate": 5.4702231931512914e-06,
+      "loss": 0.4755,
+      "step": 7628
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6035578999242772,
+      "learning_rate": 5.4691961689465565e-06,
+      "loss": 0.4976,
+      "step": 7629
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8993407079877661,
+      "learning_rate": 5.468169124769641e-06,
+      "loss": 0.6566,
+      "step": 7630
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.568348462891462,
+      "learning_rate": 5.467142060664262e-06,
+      "loss": 0.5087,
+      "step": 7631
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9491565850373038,
+      "learning_rate": 5.4661149766741385e-06,
+      "loss": 0.6116,
+      "step": 7632
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6942050574064412,
+      "learning_rate": 5.465087872842989e-06,
+      "loss": 0.5743,
+      "step": 7633
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5662075550149903,
+      "learning_rate": 5.464060749214536e-06,
+      "loss": 0.5048,
+      "step": 7634
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.878023432634221,
+      "learning_rate": 5.463033605832499e-06,
+      "loss": 0.5795,
+      "step": 7635
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6256919852216055,
+      "learning_rate": 5.4620064427406005e-06,
+      "loss": 0.5291,
+      "step": 7636
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.683820273611286,
+      "learning_rate": 5.460979259982565e-06,
+      "loss": 0.544,
+      "step": 7637
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6462524444088971,
+      "learning_rate": 5.459952057602116e-06,
+      "loss": 0.5134,
+      "step": 7638
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6712443709615155,
+      "learning_rate": 5.458924835642977e-06,
+      "loss": 0.5097,
+      "step": 7639
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.740364420120419,
+      "learning_rate": 5.457897594148876e-06,
+      "loss": 0.5932,
+      "step": 7640
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5446827282718025,
+      "learning_rate": 5.456870333163535e-06,
+      "loss": 0.4669,
+      "step": 7641
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6786201454537524,
+      "learning_rate": 5.455843052730684e-06,
+      "loss": 0.5739,
+      "step": 7642
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9191674631580304,
+      "learning_rate": 5.454815752894051e-06,
+      "loss": 0.6399,
+      "step": 7643
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7453559798957378,
+      "learning_rate": 5.4537884336973665e-06,
+      "loss": 0.5744,
+      "step": 7644
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6533767995417032,
+      "learning_rate": 5.452761095184358e-06,
+      "loss": 0.5364,
+      "step": 7645
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7773539641585071,
+      "learning_rate": 5.451733737398756e-06,
+      "loss": 0.6201,
+      "step": 7646
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6534543970475598,
+      "learning_rate": 5.4507063603842924e-06,
+      "loss": 0.536,
+      "step": 7647
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6208033437849065,
+      "learning_rate": 5.449678964184699e-06,
+      "loss": 0.5493,
+      "step": 7648
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9606060366491064,
+      "learning_rate": 5.44865154884371e-06,
+      "loss": 0.6055,
+      "step": 7649
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7554072674016378,
+      "learning_rate": 5.447624114405058e-06,
+      "loss": 0.5892,
+      "step": 7650
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7577052782835538,
+      "learning_rate": 5.446596660912478e-06,
+      "loss": 0.5999,
+      "step": 7651
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7184803457220917,
+      "learning_rate": 5.445569188409703e-06,
+      "loss": 0.6636,
+      "step": 7652
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7554933362400414,
+      "learning_rate": 5.444541696940473e-06,
+      "loss": 0.6109,
+      "step": 7653
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6525645721219876,
+      "learning_rate": 5.443514186548524e-06,
+      "loss": 0.5233,
+      "step": 7654
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8286323701669247,
+      "learning_rate": 5.442486657277593e-06,
+      "loss": 0.6284,
+      "step": 7655
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6592548567264946,
+      "learning_rate": 5.44145910917142e-06,
+      "loss": 0.6048,
+      "step": 7656
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5783620509608063,
+      "learning_rate": 5.440431542273742e-06,
+      "loss": 0.4733,
+      "step": 7657
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.636704869243039,
+      "learning_rate": 5.439403956628302e-06,
+      "loss": 0.5589,
+      "step": 7658
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6829496157119364,
+      "learning_rate": 5.438376352278839e-06,
+      "loss": 0.4903,
+      "step": 7659
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5822075259785565,
+      "learning_rate": 5.437348729269094e-06,
+      "loss": 0.5102,
+      "step": 7660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7770867537312597,
+      "learning_rate": 5.436321087642813e-06,
+      "loss": 0.5721,
+      "step": 7661
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7628500114411533,
+      "learning_rate": 5.435293427443737e-06,
+      "loss": 0.6139,
+      "step": 7662
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0242763930911243,
+      "learning_rate": 5.434265748715611e-06,
+      "loss": 0.6165,
+      "step": 7663
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8831931861429525,
+      "learning_rate": 5.4332380515021796e-06,
+      "loss": 0.5952,
+      "step": 7664
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9927954757870369,
+      "learning_rate": 5.432210335847189e-06,
+      "loss": 0.6282,
+      "step": 7665
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6387157661274618,
+      "learning_rate": 5.431182601794386e-06,
+      "loss": 0.5411,
+      "step": 7666
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8188858012671504,
+      "learning_rate": 5.430154849387515e-06,
+      "loss": 0.5674,
+      "step": 7667
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6197312203439811,
+      "learning_rate": 5.42912707867033e-06,
+      "loss": 0.4704,
+      "step": 7668
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6677575345490049,
+      "learning_rate": 5.4280992896865744e-06,
+      "loss": 0.5566,
+      "step": 7669
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5428399145339392,
+      "learning_rate": 5.427071482480001e-06,
+      "loss": 0.4867,
+      "step": 7670
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.763064482947748,
+      "learning_rate": 5.42604365709436e-06,
+      "loss": 0.5985,
+      "step": 7671
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6334282403203839,
+      "learning_rate": 5.425015813573401e-06,
+      "loss": 0.5388,
+      "step": 7672
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.865313502163088,
+      "learning_rate": 5.423987951960876e-06,
+      "loss": 0.6409,
+      "step": 7673
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6051889533901729,
+      "learning_rate": 5.42296007230054e-06,
+      "loss": 0.5532,
+      "step": 7674
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.20833480018582,
+      "learning_rate": 5.421932174636145e-06,
+      "loss": 0.6208,
+      "step": 7675
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5883651794021639,
+      "learning_rate": 5.4209042590114455e-06,
+      "loss": 0.5447,
+      "step": 7676
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7698686865402535,
+      "learning_rate": 5.419876325470197e-06,
+      "loss": 0.5948,
+      "step": 7677
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7493995078027337,
+      "learning_rate": 5.418848374056156e-06,
+      "loss": 0.625,
+      "step": 7678
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.696608713288634,
+      "learning_rate": 5.417820404813075e-06,
+      "loss": 0.593,
+      "step": 7679
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5739480096523812,
+      "learning_rate": 5.416792417784718e-06,
+      "loss": 0.4881,
+      "step": 7680
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6759667918026931,
+      "learning_rate": 5.415764413014838e-06,
+      "loss": 0.5221,
+      "step": 7681
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7777323902892049,
+      "learning_rate": 5.414736390547196e-06,
+      "loss": 0.6289,
+      "step": 7682
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7779993246484698,
+      "learning_rate": 5.4137083504255506e-06,
+      "loss": 0.656,
+      "step": 7683
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7173564828586574,
+      "learning_rate": 5.412680292693664e-06,
+      "loss": 0.6188,
+      "step": 7684
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7379808975637764,
+      "learning_rate": 5.411652217395294e-06,
+      "loss": 0.525,
+      "step": 7685
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5837568082783187,
+      "learning_rate": 5.410624124574206e-06,
+      "loss": 0.5123,
+      "step": 7686
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7511021320766276,
+      "learning_rate": 5.409596014274161e-06,
+      "loss": 0.5789,
+      "step": 7687
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8279033722935091,
+      "learning_rate": 5.408567886538922e-06,
+      "loss": 0.6342,
+      "step": 7688
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8876551816268368,
+      "learning_rate": 5.4075397414122555e-06,
+      "loss": 0.5917,
+      "step": 7689
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7098891922718354,
+      "learning_rate": 5.406511578937923e-06,
+      "loss": 0.6053,
+      "step": 7690
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8329051936661772,
+      "learning_rate": 5.405483399159691e-06,
+      "loss": 0.6319,
+      "step": 7691
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8163348430392673,
+      "learning_rate": 5.404455202121327e-06,
+      "loss": 0.6229,
+      "step": 7692
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9204914237534364,
+      "learning_rate": 5.403426987866598e-06,
+      "loss": 0.65,
+      "step": 7693
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5903937899204389,
+      "learning_rate": 5.402398756439271e-06,
+      "loss": 0.5153,
+      "step": 7694
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9063080203721176,
+      "learning_rate": 5.401370507883116e-06,
+      "loss": 0.6502,
+      "step": 7695
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7562293817119972,
+      "learning_rate": 5.400342242241899e-06,
+      "loss": 0.6432,
+      "step": 7696
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6165267075100698,
+      "learning_rate": 5.399313959559394e-06,
+      "loss": 0.519,
+      "step": 7697
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5547820862455214,
+      "learning_rate": 5.398285659879368e-06,
+      "loss": 0.5045,
+      "step": 7698
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.791158681260808,
+      "learning_rate": 5.397257343245595e-06,
+      "loss": 0.6144,
+      "step": 7699
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5458784560907478,
+      "learning_rate": 5.396229009701846e-06,
+      "loss": 0.5146,
+      "step": 7700
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7107460281972412,
+      "learning_rate": 5.395200659291895e-06,
+      "loss": 0.607,
+      "step": 7701
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9041061139278497,
+      "learning_rate": 5.394172292059514e-06,
+      "loss": 0.6222,
+      "step": 7702
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5433163520556715,
+      "learning_rate": 5.39314390804848e-06,
+      "loss": 0.4533,
+      "step": 7703
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6791522029686936,
+      "learning_rate": 5.392115507302562e-06,
+      "loss": 0.5712,
+      "step": 7704
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8935288663499319,
+      "learning_rate": 5.391087089865543e-06,
+      "loss": 0.5653,
+      "step": 7705
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6239536678147045,
+      "learning_rate": 5.390058655781196e-06,
+      "loss": 0.5159,
+      "step": 7706
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7265561882945888,
+      "learning_rate": 5.389030205093299e-06,
+      "loss": 0.5767,
+      "step": 7707
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7770451388491456,
+      "learning_rate": 5.388001737845628e-06,
+      "loss": 0.575,
+      "step": 7708
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.760367495109844,
+      "learning_rate": 5.386973254081964e-06,
+      "loss": 0.6234,
+      "step": 7709
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.773868965923522,
+      "learning_rate": 5.3859447538460855e-06,
+      "loss": 0.55,
+      "step": 7710
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5700509815885804,
+      "learning_rate": 5.384916237181771e-06,
+      "loss": 0.4986,
+      "step": 7711
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5711033557274912,
+      "learning_rate": 5.3838877041328034e-06,
+      "loss": 0.5527,
+      "step": 7712
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9728050219099901,
+      "learning_rate": 5.382859154742962e-06,
+      "loss": 0.6433,
+      "step": 7713
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.682096310741079,
+      "learning_rate": 5.3818305890560285e-06,
+      "loss": 0.5737,
+      "step": 7714
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7709356162527241,
+      "learning_rate": 5.380802007115788e-06,
+      "loss": 0.6538,
+      "step": 7715
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.701122406915414,
+      "learning_rate": 5.379773408966025e-06,
+      "loss": 0.5553,
+      "step": 7716
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9142161363093375,
+      "learning_rate": 5.37874479465052e-06,
+      "loss": 0.5796,
+      "step": 7717
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7206385256700651,
+      "learning_rate": 5.377716164213059e-06,
+      "loss": 0.5651,
+      "step": 7718
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7496814463963307,
+      "learning_rate": 5.376687517697428e-06,
+      "loss": 0.5984,
+      "step": 7719
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7629454310965097,
+      "learning_rate": 5.375658855147415e-06,
+      "loss": 0.664,
+      "step": 7720
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7155635146549625,
+      "learning_rate": 5.374630176606802e-06,
+      "loss": 0.5433,
+      "step": 7721
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7681578901605628,
+      "learning_rate": 5.373601482119381e-06,
+      "loss": 0.5762,
+      "step": 7722
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5935637803021613,
+      "learning_rate": 5.3725727717289375e-06,
+      "loss": 0.4968,
+      "step": 7723
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7929195390306502,
+      "learning_rate": 5.371544045479264e-06,
+      "loss": 0.6021,
+      "step": 7724
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7285118946070908,
+      "learning_rate": 5.370515303414146e-06,
+      "loss": 0.6269,
+      "step": 7725
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8765799796380599,
+      "learning_rate": 5.369486545577377e-06,
+      "loss": 0.6,
+      "step": 7726
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0845970718274653,
+      "learning_rate": 5.368457772012745e-06,
+      "loss": 0.5602,
+      "step": 7727
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7734004673846682,
+      "learning_rate": 5.367428982764045e-06,
+      "loss": 0.6608,
+      "step": 7728
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6317609717150813,
+      "learning_rate": 5.366400177875064e-06,
+      "loss": 0.5663,
+      "step": 7729
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5557156098948434,
+      "learning_rate": 5.3653713573896e-06,
+      "loss": 0.5377,
+      "step": 7730
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6271265655965282,
+      "learning_rate": 5.364342521351446e-06,
+      "loss": 0.5328,
+      "step": 7731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6090695624168665,
+      "learning_rate": 5.3633136698043945e-06,
+      "loss": 0.5201,
+      "step": 7732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6289026450117043,
+      "learning_rate": 5.36228480279224e-06,
+      "loss": 0.4897,
+      "step": 7733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.756977882233415,
+      "learning_rate": 5.361255920358781e-06,
+      "loss": 0.6615,
+      "step": 7734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6408538710284514,
+      "learning_rate": 5.360227022547809e-06,
+      "loss": 0.5532,
+      "step": 7735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6004730240822013,
+      "learning_rate": 5.359198109403127e-06,
+      "loss": 0.5081,
+      "step": 7736
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5559908687298513,
+      "learning_rate": 5.358169180968527e-06,
+      "loss": 0.4976,
+      "step": 7737
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6404008594146845,
+      "learning_rate": 5.357140237287811e-06,
+      "loss": 0.5139,
+      "step": 7738
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7470665460472893,
+      "learning_rate": 5.356111278404777e-06,
+      "loss": 0.626,
+      "step": 7739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5834475636775669,
+      "learning_rate": 5.355082304363221e-06,
+      "loss": 0.4851,
+      "step": 7740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5896010140170032,
+      "learning_rate": 5.354053315206947e-06,
+      "loss": 0.4941,
+      "step": 7741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0499949451309503,
+      "learning_rate": 5.353024310979757e-06,
+      "loss": 0.6522,
+      "step": 7742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5054924310076513,
+      "learning_rate": 5.351995291725448e-06,
+      "loss": 0.4554,
+      "step": 7743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6677973829308487,
+      "learning_rate": 5.350966257487825e-06,
+      "loss": 0.5491,
+      "step": 7744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7075294731007261,
+      "learning_rate": 5.34993720831069e-06,
+      "loss": 0.5112,
+      "step": 7745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9745288610078423,
+      "learning_rate": 5.348908144237846e-06,
+      "loss": 0.6473,
+      "step": 7746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5622815945261704,
+      "learning_rate": 5.347879065313096e-06,
+      "loss": 0.4875,
+      "step": 7747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6692257317838569,
+      "learning_rate": 5.346849971580248e-06,
+      "loss": 0.5901,
+      "step": 7748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6786583804214478,
+      "learning_rate": 5.345820863083105e-06,
+      "loss": 0.5365,
+      "step": 7749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6105684261085144,
+      "learning_rate": 5.344791739865471e-06,
+      "loss": 0.5546,
+      "step": 7750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8076388076080914,
+      "learning_rate": 5.343762601971156e-06,
+      "loss": 0.6665,
+      "step": 7751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8570977591880408,
+      "learning_rate": 5.342733449443965e-06,
+      "loss": 0.6219,
+      "step": 7752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6370562571564155,
+      "learning_rate": 5.341704282327705e-06,
+      "loss": 0.5322,
+      "step": 7753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6260290500439879,
+      "learning_rate": 5.340675100666188e-06,
+      "loss": 0.5135,
+      "step": 7754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5904676112175424,
+      "learning_rate": 5.33964590450322e-06,
+      "loss": 0.4997,
+      "step": 7755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7451974252699439,
+      "learning_rate": 5.338616693882611e-06,
+      "loss": 0.5632,
+      "step": 7756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7192540204047387,
+      "learning_rate": 5.337587468848171e-06,
+      "loss": 0.5441,
+      "step": 7757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.622893404555686,
+      "learning_rate": 5.336558229443712e-06,
+      "loss": 0.547,
+      "step": 7758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6279810475129116,
+      "learning_rate": 5.335528975713045e-06,
+      "loss": 0.4898,
+      "step": 7759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7152529339530527,
+      "learning_rate": 5.33449970769998e-06,
+      "loss": 0.5411,
+      "step": 7760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7052606351232354,
+      "learning_rate": 5.333470425448332e-06,
+      "loss": 0.603,
+      "step": 7761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5823979446585354,
+      "learning_rate": 5.332441129001914e-06,
+      "loss": 0.5626,
+      "step": 7762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7608170523306291,
+      "learning_rate": 5.331411818404539e-06,
+      "loss": 0.5586,
+      "step": 7763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6128208302282596,
+      "learning_rate": 5.3303824937000225e-06,
+      "loss": 0.5629,
+      "step": 7764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7164929974776274,
+      "learning_rate": 5.329353154932179e-06,
+      "loss": 0.5765,
+      "step": 7765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7713169793183182,
+      "learning_rate": 5.328323802144822e-06,
+      "loss": 0.5695,
+      "step": 7766
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9265304642052163,
+      "learning_rate": 5.327294435381772e-06,
+      "loss": 0.6818,
+      "step": 7767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.825508822067408,
+      "learning_rate": 5.326265054686844e-06,
+      "loss": 0.5846,
+      "step": 7768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.628665724668007,
+      "learning_rate": 5.325235660103856e-06,
+      "loss": 0.5755,
+      "step": 7769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.66366246560793,
+      "learning_rate": 5.324206251676623e-06,
+      "loss": 0.5227,
+      "step": 7770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6761650849348263,
+      "learning_rate": 5.323176829448967e-06,
+      "loss": 0.5173,
+      "step": 7771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7199382572637044,
+      "learning_rate": 5.322147393464706e-06,
+      "loss": 0.5702,
+      "step": 7772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5887790917193235,
+      "learning_rate": 5.321117943767661e-06,
+      "loss": 0.5038,
+      "step": 7773
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.031051151846374,
+      "learning_rate": 5.320088480401649e-06,
+      "loss": 0.6142,
+      "step": 7774
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7283028926059821,
+      "learning_rate": 5.319059003410496e-06,
+      "loss": 0.5782,
+      "step": 7775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7614281547964779,
+      "learning_rate": 5.318029512838018e-06,
+      "loss": 0.6233,
+      "step": 7776
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7501909888112255,
+      "learning_rate": 5.317000008728042e-06,
+      "loss": 0.5937,
+      "step": 7777
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.737652573312114,
+      "learning_rate": 5.315970491124387e-06,
+      "loss": 0.5731,
+      "step": 7778
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7259415196763576,
+      "learning_rate": 5.314940960070879e-06,
+      "loss": 0.5387,
+      "step": 7779
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5689656767728895,
+      "learning_rate": 5.313911415611341e-06,
+      "loss": 0.5163,
+      "step": 7780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2708990744823239,
+      "learning_rate": 5.312881857789596e-06,
+      "loss": 0.6402,
+      "step": 7781
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6190638294954918,
+      "learning_rate": 5.31185228664947e-06,
+      "loss": 0.5644,
+      "step": 7782
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6110976970031593,
+      "learning_rate": 5.31082270223479e-06,
+      "loss": 0.5053,
+      "step": 7783
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5809708065534906,
+      "learning_rate": 5.309793104589379e-06,
+      "loss": 0.4796,
+      "step": 7784
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5963783041967836,
+      "learning_rate": 5.308763493757067e-06,
+      "loss": 0.517,
+      "step": 7785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7553283451643742,
+      "learning_rate": 5.30773386978168e-06,
+      "loss": 0.5929,
+      "step": 7786
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.812690534696544,
+      "learning_rate": 5.306704232707045e-06,
+      "loss": 0.6016,
+      "step": 7787
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6015041572317619,
+      "learning_rate": 5.305674582576991e-06,
+      "loss": 0.4999,
+      "step": 7788
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.089798022059888,
+      "learning_rate": 5.304644919435347e-06,
+      "loss": 0.6408,
+      "step": 7789
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8056609174808825,
+      "learning_rate": 5.303615243325942e-06,
+      "loss": 0.5716,
+      "step": 7790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8010464937966264,
+      "learning_rate": 5.302585554292606e-06,
+      "loss": 0.5956,
+      "step": 7791
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6617620654217385,
+      "learning_rate": 5.30155585237917e-06,
+      "loss": 0.5772,
+      "step": 7792
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5711171743383584,
+      "learning_rate": 5.300526137629465e-06,
+      "loss": 0.5067,
+      "step": 7793
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5805484877470243,
+      "learning_rate": 5.2994964100873225e-06,
+      "loss": 0.5578,
+      "step": 7794
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5776063981112649,
+      "learning_rate": 5.2984666697965755e-06,
+      "loss": 0.4718,
+      "step": 7795
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7996069024995108,
+      "learning_rate": 5.297436916801057e-06,
+      "loss": 0.6427,
+      "step": 7796
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7790259147680327,
+      "learning_rate": 5.296407151144597e-06,
+      "loss": 0.631,
+      "step": 7797
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6982858058000818,
+      "learning_rate": 5.295377372871033e-06,
+      "loss": 0.5038,
+      "step": 7798
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6696859042299708,
+      "learning_rate": 5.2943475820241975e-06,
+      "loss": 0.5594,
+      "step": 7799
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9531840280237015,
+      "learning_rate": 5.293317778647927e-06,
+      "loss": 0.6101,
+      "step": 7800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6875684693717123,
+      "learning_rate": 5.292287962786055e-06,
+      "loss": 0.5306,
+      "step": 7801
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.713596805921403,
+      "learning_rate": 5.29125813448242e-06,
+      "loss": 0.5671,
+      "step": 7802
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5320728376249004,
+      "learning_rate": 5.290228293780855e-06,
+      "loss": 0.4567,
+      "step": 7803
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.774006308745748,
+      "learning_rate": 5.2891984407252e-06,
+      "loss": 0.6245,
+      "step": 7804
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5712094596877593,
+      "learning_rate": 5.2881685753592915e-06,
+      "loss": 0.4808,
+      "step": 7805
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6623986780775507,
+      "learning_rate": 5.2871386977269675e-06,
+      "loss": 0.5595,
+      "step": 7806
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7005396887928569,
+      "learning_rate": 5.286108807872068e-06,
+      "loss": 0.5233,
+      "step": 7807
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7504345644670651,
+      "learning_rate": 5.28507890583843e-06,
+      "loss": 0.5768,
+      "step": 7808
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7364170525226253,
+      "learning_rate": 5.284048991669892e-06,
+      "loss": 0.5746,
+      "step": 7809
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9293607265773179,
+      "learning_rate": 5.283019065410298e-06,
+      "loss": 0.579,
+      "step": 7810
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8804876851137308,
+      "learning_rate": 5.281989127103486e-06,
+      "loss": 0.6288,
+      "step": 7811
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5648661112293749,
+      "learning_rate": 5.280959176793299e-06,
+      "loss": 0.5166,
+      "step": 7812
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.641143481452317,
+      "learning_rate": 5.279929214523577e-06,
+      "loss": 0.5385,
+      "step": 7813
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7492285420082699,
+      "learning_rate": 5.278899240338164e-06,
+      "loss": 0.5846,
+      "step": 7814
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0233634774609417,
+      "learning_rate": 5.277869254280899e-06,
+      "loss": 0.5459,
+      "step": 7815
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6057256830890536,
+      "learning_rate": 5.27683925639563e-06,
+      "loss": 0.543,
+      "step": 7816
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5877385092786994,
+      "learning_rate": 5.275809246726198e-06,
+      "loss": 0.5288,
+      "step": 7817
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9096373109851044,
+      "learning_rate": 5.2747792253164475e-06,
+      "loss": 0.6386,
+      "step": 7818
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.628624302150182,
+      "learning_rate": 5.273749192210223e-06,
+      "loss": 0.5669,
+      "step": 7819
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5481796449252005,
+      "learning_rate": 5.272719147451372e-06,
+      "loss": 0.519,
+      "step": 7820
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0152923342247557,
+      "learning_rate": 5.271689091083737e-06,
+      "loss": 0.6053,
+      "step": 7821
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0712457774012485,
+      "learning_rate": 5.270659023151164e-06,
+      "loss": 0.614,
+      "step": 7822
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.688658659315264,
+      "learning_rate": 5.269628943697504e-06,
+      "loss": 0.5533,
+      "step": 7823
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6756675346791414,
+      "learning_rate": 5.2685988527666e-06,
+      "loss": 0.5231,
+      "step": 7824
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5546344409186675,
+      "learning_rate": 5.267568750402302e-06,
+      "loss": 0.4822,
+      "step": 7825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7301359025716571,
+      "learning_rate": 5.266538636648457e-06,
+      "loss": 0.5981,
+      "step": 7826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6749830205492546,
+      "learning_rate": 5.265508511548914e-06,
+      "loss": 0.5418,
+      "step": 7827
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7416448440544625,
+      "learning_rate": 5.26447837514752e-06,
+      "loss": 0.5355,
+      "step": 7828
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5509875832044191,
+      "learning_rate": 5.263448227488129e-06,
+      "loss": 0.5682,
+      "step": 7829
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.07279499002777,
+      "learning_rate": 5.26241806861459e-06,
+      "loss": 0.5951,
+      "step": 7830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5859297692953886,
+      "learning_rate": 5.26138789857075e-06,
+      "loss": 0.5068,
+      "step": 7831
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7927452168983924,
+      "learning_rate": 5.260357717400464e-06,
+      "loss": 0.6395,
+      "step": 7832
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6918959233784057,
+      "learning_rate": 5.2593275251475815e-06,
+      "loss": 0.5167,
+      "step": 7833
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5879944535026977,
+      "learning_rate": 5.258297321855955e-06,
+      "loss": 0.555,
+      "step": 7834
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7328910998645103,
+      "learning_rate": 5.257267107569437e-06,
+      "loss": 0.5177,
+      "step": 7835
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6526112273473237,
+      "learning_rate": 5.25623688233188e-06,
+      "loss": 0.4904,
+      "step": 7836
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.780447083648694,
+      "learning_rate": 5.255206646187137e-06,
+      "loss": 0.6327,
+      "step": 7837
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.603639518509909,
+      "learning_rate": 5.254176399179063e-06,
+      "loss": 0.5729,
+      "step": 7838
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8902777452328379,
+      "learning_rate": 5.253146141351513e-06,
+      "loss": 0.5984,
+      "step": 7839
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6154693779911932,
+      "learning_rate": 5.252115872748339e-06,
+      "loss": 0.5397,
+      "step": 7840
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6534417590394038,
+      "learning_rate": 5.2510855934134e-06,
+      "loss": 0.5333,
+      "step": 7841
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6595574881738436,
+      "learning_rate": 5.2500553033905475e-06,
+      "loss": 0.5312,
+      "step": 7842
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6746130595493086,
+      "learning_rate": 5.249025002723641e-06,
+      "loss": 0.5419,
+      "step": 7843
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6270313439856194,
+      "learning_rate": 5.247994691456536e-06,
+      "loss": 0.5279,
+      "step": 7844
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.775804032349742,
+      "learning_rate": 5.24696436963309e-06,
+      "loss": 0.5997,
+      "step": 7845
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.832411540395696,
+      "learning_rate": 5.2459340372971575e-06,
+      "loss": 0.5826,
+      "step": 7846
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6519801525397707,
+      "learning_rate": 5.244903694492601e-06,
+      "loss": 0.5549,
+      "step": 7847
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6134851027548354,
+      "learning_rate": 5.243873341263277e-06,
+      "loss": 0.4926,
+      "step": 7848
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7119932640697326,
+      "learning_rate": 5.242842977653043e-06,
+      "loss": 0.5691,
+      "step": 7849
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2222442909906404,
+      "learning_rate": 5.24181260370576e-06,
+      "loss": 0.5123,
+      "step": 7850
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5758065634254151,
+      "learning_rate": 5.240782219465288e-06,
+      "loss": 0.4688,
+      "step": 7851
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6321561441497726,
+      "learning_rate": 5.239751824975486e-06,
+      "loss": 0.462,
+      "step": 7852
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.558115720826898,
+      "learning_rate": 5.238721420280214e-06,
+      "loss": 0.4966,
+      "step": 7853
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5588446667874375,
+      "learning_rate": 5.2376910054233345e-06,
+      "loss": 0.4864,
+      "step": 7854
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5867568047909654,
+      "learning_rate": 5.236660580448708e-06,
+      "loss": 0.4867,
+      "step": 7855
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5976199452892107,
+      "learning_rate": 5.2356301454001975e-06,
+      "loss": 0.4574,
+      "step": 7856
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.57499500957606,
+      "learning_rate": 5.234599700321665e-06,
+      "loss": 0.4515,
+      "step": 7857
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6237006656153087,
+      "learning_rate": 5.233569245256972e-06,
+      "loss": 0.4501,
+      "step": 7858
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6916216831897695,
+      "learning_rate": 5.232538780249983e-06,
+      "loss": 0.4787,
+      "step": 7859
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6085300666010587,
+      "learning_rate": 5.23150830534456e-06,
+      "loss": 0.4869,
+      "step": 7860
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8276365431073447,
+      "learning_rate": 5.230477820584571e-06,
+      "loss": 0.5197,
+      "step": 7861
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5487516361202578,
+      "learning_rate": 5.2294473260138755e-06,
+      "loss": 0.4347,
+      "step": 7862
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5517042728327083,
+      "learning_rate": 5.228416821676342e-06,
+      "loss": 0.4777,
+      "step": 7863
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6183710365216042,
+      "learning_rate": 5.227386307615834e-06,
+      "loss": 0.5175,
+      "step": 7864
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6792680583163542,
+      "learning_rate": 5.226355783876216e-06,
+      "loss": 0.5101,
+      "step": 7865
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.528404102546661,
+      "learning_rate": 5.225325250501356e-06,
+      "loss": 0.4194,
+      "step": 7866
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5584978234264941,
+      "learning_rate": 5.224294707535121e-06,
+      "loss": 0.4969,
+      "step": 7867
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6369438303944066,
+      "learning_rate": 5.223264155021377e-06,
+      "loss": 0.4655,
+      "step": 7868
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5845717616783592,
+      "learning_rate": 5.222233593003991e-06,
+      "loss": 0.4757,
+      "step": 7869
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5806164828941629,
+      "learning_rate": 5.2212030215268316e-06,
+      "loss": 0.4334,
+      "step": 7870
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6175428783073746,
+      "learning_rate": 5.220172440633765e-06,
+      "loss": 0.4572,
+      "step": 7871
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6714088486505605,
+      "learning_rate": 5.219141850368663e-06,
+      "loss": 0.5005,
+      "step": 7872
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6064130224923959,
+      "learning_rate": 5.218111250775392e-06,
+      "loss": 0.4589,
+      "step": 7873
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5881835032905863,
+      "learning_rate": 5.217080641897822e-06,
+      "loss": 0.4547,
+      "step": 7874
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5449062120445409,
+      "learning_rate": 5.216050023779823e-06,
+      "loss": 0.4079,
+      "step": 7875
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5739006332696504,
+      "learning_rate": 5.215019396465265e-06,
+      "loss": 0.466,
+      "step": 7876
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7404029981314925,
+      "learning_rate": 5.2139887599980165e-06,
+      "loss": 0.5023,
+      "step": 7877
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5806108049422828,
+      "learning_rate": 5.212958114421952e-06,
+      "loss": 0.4803,
+      "step": 7878
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6495553722544288,
+      "learning_rate": 5.211927459780941e-06,
+      "loss": 0.4933,
+      "step": 7879
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7496193297608316,
+      "learning_rate": 5.210896796118856e-06,
+      "loss": 0.5363,
+      "step": 7880
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7422031504574214,
+      "learning_rate": 5.209866123479568e-06,
+      "loss": 0.5495,
+      "step": 7881
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2325166974738717,
+      "learning_rate": 5.208835441906949e-06,
+      "loss": 0.5266,
+      "step": 7882
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5955422006311042,
+      "learning_rate": 5.207804751444873e-06,
+      "loss": 0.4479,
+      "step": 7883
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6866498561064781,
+      "learning_rate": 5.206774052137211e-06,
+      "loss": 0.5006,
+      "step": 7884
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7055277263865362,
+      "learning_rate": 5.205743344027841e-06,
+      "loss": 0.5079,
+      "step": 7885
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7765973950804936,
+      "learning_rate": 5.204712627160633e-06,
+      "loss": 0.5159,
+      "step": 7886
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8552237561245664,
+      "learning_rate": 5.203681901579463e-06,
+      "loss": 0.5298,
+      "step": 7887
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7650165748175736,
+      "learning_rate": 5.2026511673282055e-06,
+      "loss": 0.5489,
+      "step": 7888
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7784502397602112,
+      "learning_rate": 5.201620424450735e-06,
+      "loss": 0.5351,
+      "step": 7889
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.717783693811607,
+      "learning_rate": 5.2005896729909265e-06,
+      "loss": 0.5285,
+      "step": 7890
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6732178759992189,
+      "learning_rate": 5.1995589129926584e-06,
+      "loss": 0.5296,
+      "step": 7891
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7143077838780056,
+      "learning_rate": 5.1985281444998035e-06,
+      "loss": 0.5036,
+      "step": 7892
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5784084365810707,
+      "learning_rate": 5.1974973675562415e-06,
+      "loss": 0.4775,
+      "step": 7893
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7220102673155449,
+      "learning_rate": 5.196466582205847e-06,
+      "loss": 0.5045,
+      "step": 7894
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6974815958785543,
+      "learning_rate": 5.195435788492498e-06,
+      "loss": 0.5339,
+      "step": 7895
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.661259867635626,
+      "learning_rate": 5.194404986460072e-06,
+      "loss": 0.5089,
+      "step": 7896
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6067626326698959,
+      "learning_rate": 5.193374176152447e-06,
+      "loss": 0.4607,
+      "step": 7897
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5788263325410276,
+      "learning_rate": 5.192343357613501e-06,
+      "loss": 0.4585,
+      "step": 7898
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5907338388077803,
+      "learning_rate": 5.191312530887111e-06,
+      "loss": 0.4082,
+      "step": 7899
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6836441289086129,
+      "learning_rate": 5.190281696017161e-06,
+      "loss": 0.4782,
+      "step": 7900
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7343299955143782,
+      "learning_rate": 5.189250853047526e-06,
+      "loss": 0.4759,
+      "step": 7901
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7058577594577498,
+      "learning_rate": 5.1882200020220865e-06,
+      "loss": 0.5097,
+      "step": 7902
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5836534081409047,
+      "learning_rate": 5.187189142984724e-06,
+      "loss": 0.4643,
+      "step": 7903
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5992547612054283,
+      "learning_rate": 5.186158275979317e-06,
+      "loss": 0.5217,
+      "step": 7904
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7516662410073572,
+      "learning_rate": 5.185127401049747e-06,
+      "loss": 0.5022,
+      "step": 7905
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6707635254400446,
+      "learning_rate": 5.184096518239896e-06,
+      "loss": 0.5076,
+      "step": 7906
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9783763335461778,
+      "learning_rate": 5.183065627593644e-06,
+      "loss": 0.5246,
+      "step": 7907
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6263104276982511,
+      "learning_rate": 5.182034729154873e-06,
+      "loss": 0.4564,
+      "step": 7908
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8231500856341674,
+      "learning_rate": 5.1810038229674655e-06,
+      "loss": 0.499,
+      "step": 7909
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6786524373082691,
+      "learning_rate": 5.179972909075304e-06,
+      "loss": 0.5495,
+      "step": 7910
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8591074376454022,
+      "learning_rate": 5.17894198752227e-06,
+      "loss": 0.6068,
+      "step": 7911
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5978885680276532,
+      "learning_rate": 5.177911058352249e-06,
+      "loss": 0.4423,
+      "step": 7912
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5953491452164323,
+      "learning_rate": 5.1768801216091214e-06,
+      "loss": 0.4332,
+      "step": 7913
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6073291604471678,
+      "learning_rate": 5.175849177336772e-06,
+      "loss": 0.445,
+      "step": 7914
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5661976490389823,
+      "learning_rate": 5.1748182255790854e-06,
+      "loss": 0.4511,
+      "step": 7915
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7594191405537379,
+      "learning_rate": 5.173787266379946e-06,
+      "loss": 0.4677,
+      "step": 7916
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5494028434496976,
+      "learning_rate": 5.1727562997832385e-06,
+      "loss": 0.4489,
+      "step": 7917
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5562907041862178,
+      "learning_rate": 5.171725325832846e-06,
+      "loss": 0.4004,
+      "step": 7918
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6547373523823373,
+      "learning_rate": 5.170694344572656e-06,
+      "loss": 0.5031,
+      "step": 7919
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.716585560276361,
+      "learning_rate": 5.169663356046554e-06,
+      "loss": 0.5387,
+      "step": 7920
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6218172132755906,
+      "learning_rate": 5.168632360298422e-06,
+      "loss": 0.4946,
+      "step": 7921
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6218170480458325,
+      "learning_rate": 5.167601357372152e-06,
+      "loss": 0.4442,
+      "step": 7922
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6189001511113672,
+      "learning_rate": 5.1665703473116276e-06,
+      "loss": 0.4465,
+      "step": 7923
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5814782125395176,
+      "learning_rate": 5.165539330160736e-06,
+      "loss": 0.4239,
+      "step": 7924
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6314575397965841,
+      "learning_rate": 5.164508305963363e-06,
+      "loss": 0.4803,
+      "step": 7925
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5776419503010333,
+      "learning_rate": 5.163477274763398e-06,
+      "loss": 0.4772,
+      "step": 7926
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6093649946999745,
+      "learning_rate": 5.162446236604727e-06,
+      "loss": 0.4915,
+      "step": 7927
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6529769421281877,
+      "learning_rate": 5.161415191531238e-06,
+      "loss": 0.48,
+      "step": 7928
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6524811141356149,
+      "learning_rate": 5.160384139586823e-06,
+      "loss": 0.4926,
+      "step": 7929
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7418814915815949,
+      "learning_rate": 5.159353080815366e-06,
+      "loss": 0.548,
+      "step": 7930
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.880659114316506,
+      "learning_rate": 5.1583220152607576e-06,
+      "loss": 0.4367,
+      "step": 7931
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7589214795491664,
+      "learning_rate": 5.157290942966887e-06,
+      "loss": 0.4815,
+      "step": 7932
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7780801338528642,
+      "learning_rate": 5.156259863977642e-06,
+      "loss": 0.5576,
+      "step": 7933
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6991793599579742,
+      "learning_rate": 5.155228778336916e-06,
+      "loss": 0.475,
+      "step": 7934
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6366171494469306,
+      "learning_rate": 5.154197686088597e-06,
+      "loss": 0.483,
+      "step": 7935
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8350875705969815,
+      "learning_rate": 5.153166587276575e-06,
+      "loss": 0.5303,
+      "step": 7936
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6537874640664983,
+      "learning_rate": 5.15213548194474e-06,
+      "loss": 0.4723,
+      "step": 7937
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7102780936488094,
+      "learning_rate": 5.151104370136985e-06,
+      "loss": 0.5038,
+      "step": 7938
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6163822227108454,
+      "learning_rate": 5.150073251897197e-06,
+      "loss": 0.4586,
+      "step": 7939
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8634455972841117,
+      "learning_rate": 5.149042127269273e-06,
+      "loss": 0.4737,
+      "step": 7940
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6060153083147701,
+      "learning_rate": 5.148010996297101e-06,
+      "loss": 0.4322,
+      "step": 7941
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6140990846526703,
+      "learning_rate": 5.146979859024575e-06,
+      "loss": 0.4033,
+      "step": 7942
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6012467827600491,
+      "learning_rate": 5.1459487154955845e-06,
+      "loss": 0.4946,
+      "step": 7943
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.636700584098543,
+      "learning_rate": 5.144917565754024e-06,
+      "loss": 0.4964,
+      "step": 7944
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7500634117767534,
+      "learning_rate": 5.143886409843787e-06,
+      "loss": 0.493,
+      "step": 7945
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5440606391210333,
+      "learning_rate": 5.142855247808763e-06,
+      "loss": 0.4259,
+      "step": 7946
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6452696746661204,
+      "learning_rate": 5.141824079692849e-06,
+      "loss": 0.4542,
+      "step": 7947
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6072678091343034,
+      "learning_rate": 5.140792905539936e-06,
+      "loss": 0.4354,
+      "step": 7948
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.632203985306397,
+      "learning_rate": 5.1397617253939214e-06,
+      "loss": 0.4692,
+      "step": 7949
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6658873492179695,
+      "learning_rate": 5.1387305392986955e-06,
+      "loss": 0.5216,
+      "step": 7950
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7683551914337232,
+      "learning_rate": 5.137699347298153e-06,
+      "loss": 0.5434,
+      "step": 7951
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5920245185048284,
+      "learning_rate": 5.136668149436189e-06,
+      "loss": 0.4606,
+      "step": 7952
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7175111090016661,
+      "learning_rate": 5.1356369457567e-06,
+      "loss": 0.521,
+      "step": 7953
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5766744624633092,
+      "learning_rate": 5.1346057363035796e-06,
+      "loss": 0.4558,
+      "step": 7954
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6628743339456539,
+      "learning_rate": 5.133574521120723e-06,
+      "loss": 0.432,
+      "step": 7955
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6042426929200225,
+      "learning_rate": 5.132543300252026e-06,
+      "loss": 0.4063,
+      "step": 7956
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6178481322734461,
+      "learning_rate": 5.131512073741383e-06,
+      "loss": 0.4723,
+      "step": 7957
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7220445758293353,
+      "learning_rate": 5.1304808416326935e-06,
+      "loss": 0.4703,
+      "step": 7958
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5529690856768076,
+      "learning_rate": 5.12944960396985e-06,
+      "loss": 0.4275,
+      "step": 7959
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5661459335550141,
+      "learning_rate": 5.128418360796751e-06,
+      "loss": 0.4271,
+      "step": 7960
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7664397795751051,
+      "learning_rate": 5.1273871121572916e-06,
+      "loss": 0.5452,
+      "step": 7961
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7308487637051548,
+      "learning_rate": 5.126355858095371e-06,
+      "loss": 0.5179,
+      "step": 7962
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8226834474919791,
+      "learning_rate": 5.125324598654885e-06,
+      "loss": 0.5422,
+      "step": 7963
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5945051764222167,
+      "learning_rate": 5.1242933338797315e-06,
+      "loss": 0.4644,
+      "step": 7964
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6473159898242234,
+      "learning_rate": 5.123262063813809e-06,
+      "loss": 0.469,
+      "step": 7965
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7744216021454134,
+      "learning_rate": 5.1222307885010125e-06,
+      "loss": 0.5292,
+      "step": 7966
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.644001575144423,
+      "learning_rate": 5.121199507985243e-06,
+      "loss": 0.5379,
+      "step": 7967
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8232099811801318,
+      "learning_rate": 5.120168222310398e-06,
+      "loss": 0.4931,
+      "step": 7968
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6478201516002157,
+      "learning_rate": 5.119136931520374e-06,
+      "loss": 0.4849,
+      "step": 7969
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6098348884653333,
+      "learning_rate": 5.118105635659072e-06,
+      "loss": 0.4889,
+      "step": 7970
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7094914566743494,
+      "learning_rate": 5.1170743347703925e-06,
+      "loss": 0.5458,
+      "step": 7971
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5057615961880692,
+      "learning_rate": 5.116043028898231e-06,
+      "loss": 0.4074,
+      "step": 7972
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6252288079856869,
+      "learning_rate": 5.1150117180864885e-06,
+      "loss": 0.4581,
+      "step": 7973
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7702721521291179,
+      "learning_rate": 5.113980402379066e-06,
+      "loss": 0.4721,
+      "step": 7974
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6409602156185231,
+      "learning_rate": 5.112949081819861e-06,
+      "loss": 0.4741,
+      "step": 7975
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7036416261389531,
+      "learning_rate": 5.1119177564527744e-06,
+      "loss": 0.5164,
+      "step": 7976
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6132219639646236,
+      "learning_rate": 5.110886426321706e-06,
+      "loss": 0.48,
+      "step": 7977
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6490586193228038,
+      "learning_rate": 5.109855091470558e-06,
+      "loss": 0.4333,
+      "step": 7978
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6223689032668894,
+      "learning_rate": 5.108823751943229e-06,
+      "loss": 0.4675,
+      "step": 7979
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7359589583358829,
+      "learning_rate": 5.107792407783621e-06,
+      "loss": 0.5063,
+      "step": 7980
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7285229381058957,
+      "learning_rate": 5.106761059035635e-06,
+      "loss": 0.4765,
+      "step": 7981
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8333369552148613,
+      "learning_rate": 5.105729705743172e-06,
+      "loss": 0.5086,
+      "step": 7982
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6623259111464884,
+      "learning_rate": 5.104698347950133e-06,
+      "loss": 0.471,
+      "step": 7983
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6391896574243159,
+      "learning_rate": 5.103666985700419e-06,
+      "loss": 0.3967,
+      "step": 7984
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5695048795154143,
+      "learning_rate": 5.102635619037933e-06,
+      "loss": 0.4193,
+      "step": 7985
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6258258704388905,
+      "learning_rate": 5.101604248006578e-06,
+      "loss": 0.4494,
+      "step": 7986
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5364360820448856,
+      "learning_rate": 5.100572872650253e-06,
+      "loss": 0.4197,
+      "step": 7987
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5681300376540526,
+      "learning_rate": 5.099541493012864e-06,
+      "loss": 0.4309,
+      "step": 7988
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6438331062224097,
+      "learning_rate": 5.098510109138311e-06,
+      "loss": 0.4402,
+      "step": 7989
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7032718223863288,
+      "learning_rate": 5.097478721070497e-06,
+      "loss": 0.442,
+      "step": 7990
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2396032828535386,
+      "learning_rate": 5.096447328853325e-06,
+      "loss": 0.5068,
+      "step": 7991
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6757881311741756,
+      "learning_rate": 5.095415932530699e-06,
+      "loss": 0.4868,
+      "step": 7992
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6961110881329763,
+      "learning_rate": 5.094384532146522e-06,
+      "loss": 0.5216,
+      "step": 7993
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7916969372080543,
+      "learning_rate": 5.093353127744698e-06,
+      "loss": 0.4935,
+      "step": 7994
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7472611838981202,
+      "learning_rate": 5.092321719369127e-06,
+      "loss": 0.4576,
+      "step": 7995
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7628731599237817,
+      "learning_rate": 5.091290307063718e-06,
+      "loss": 0.5065,
+      "step": 7996
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5843069535493849,
+      "learning_rate": 5.09025889087237e-06,
+      "loss": 0.4872,
+      "step": 7997
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8597133104545779,
+      "learning_rate": 5.0892274708389915e-06,
+      "loss": 0.5905,
+      "step": 7998
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8233981790207843,
+      "learning_rate": 5.088196047007484e-06,
+      "loss": 0.5543,
+      "step": 7999
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8092275421862285,
+      "learning_rate": 5.087164619421753e-06,
+      "loss": 0.5465,
+      "step": 8000
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.1104018329533105,
+      "learning_rate": 5.0861331881257005e-06,
+      "loss": 0.5639,
+      "step": 8001
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7982239717209036,
+      "learning_rate": 5.085101753163235e-06,
+      "loss": 0.5469,
+      "step": 8002
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7441514512711345,
+      "learning_rate": 5.084070314578261e-06,
+      "loss": 0.5199,
+      "step": 8003
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.63587168828236,
+      "learning_rate": 5.083038872414681e-06,
+      "loss": 0.4655,
+      "step": 8004
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.505826891907447,
+      "learning_rate": 5.082007426716402e-06,
+      "loss": 0.4379,
+      "step": 8005
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5715545494586659,
+      "learning_rate": 5.080975977527329e-06,
+      "loss": 0.4634,
+      "step": 8006
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6008932968992711,
+      "learning_rate": 5.079944524891367e-06,
+      "loss": 0.4287,
+      "step": 8007
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5582738378860898,
+      "learning_rate": 5.078913068852421e-06,
+      "loss": 0.4634,
+      "step": 8008
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6942054465762226,
+      "learning_rate": 5.077881609454399e-06,
+      "loss": 0.4845,
+      "step": 8009
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.845694199068934,
+      "learning_rate": 5.076850146741207e-06,
+      "loss": 0.5073,
+      "step": 8010
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6477070260379001,
+      "learning_rate": 5.075818680756749e-06,
+      "loss": 0.4626,
+      "step": 8011
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6164803333648873,
+      "learning_rate": 5.074787211544931e-06,
+      "loss": 0.4639,
+      "step": 8012
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6111837475395059,
+      "learning_rate": 5.0737557391496615e-06,
+      "loss": 0.4568,
+      "step": 8013
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6199208845182491,
+      "learning_rate": 5.0727242636148445e-06,
+      "loss": 0.5111,
+      "step": 8014
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6862575354801447,
+      "learning_rate": 5.071692784984389e-06,
+      "loss": 0.4912,
+      "step": 8015
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.9000295164859495,
+      "learning_rate": 5.070661303302201e-06,
+      "loss": 0.4288,
+      "step": 8016
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8527616750432485,
+      "learning_rate": 5.069629818612186e-06,
+      "loss": 0.5453,
+      "step": 8017
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5989249596671377,
+      "learning_rate": 5.068598330958253e-06,
+      "loss": 0.4164,
+      "step": 8018
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6319806794869235,
+      "learning_rate": 5.067566840384309e-06,
+      "loss": 0.4262,
+      "step": 8019
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5789886376819435,
+      "learning_rate": 5.066535346934259e-06,
+      "loss": 0.4989,
+      "step": 8020
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5957754257489064,
+      "learning_rate": 5.065503850652014e-06,
+      "loss": 0.4031,
+      "step": 8021
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6239028935529184,
+      "learning_rate": 5.064472351581478e-06,
+      "loss": 0.5024,
+      "step": 8022
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8934761990814896,
+      "learning_rate": 5.063440849766559e-06,
+      "loss": 0.5414,
+      "step": 8023
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5852631394463695,
+      "learning_rate": 5.062409345251167e-06,
+      "loss": 0.4494,
+      "step": 8024
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6537072578740031,
+      "learning_rate": 5.0613778380792075e-06,
+      "loss": 0.433,
+      "step": 8025
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5423354152096477,
+      "learning_rate": 5.06034632829459e-06,
+      "loss": 0.4414,
+      "step": 8026
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6675241775017613,
+      "learning_rate": 5.059314815941224e-06,
+      "loss": 0.4474,
+      "step": 8027
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9803082593937613,
+      "learning_rate": 5.058283301063014e-06,
+      "loss": 0.5153,
+      "step": 8028
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7927231404575348,
+      "learning_rate": 5.057251783703871e-06,
+      "loss": 0.4663,
+      "step": 8029
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5923349367851576,
+      "learning_rate": 5.056220263907702e-06,
+      "loss": 0.4421,
+      "step": 8030
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5537083884691775,
+      "learning_rate": 5.055188741718416e-06,
+      "loss": 0.4337,
+      "step": 8031
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.594545552648897,
+      "learning_rate": 5.054157217179922e-06,
+      "loss": 0.4941,
+      "step": 8032
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8084866070847276,
+      "learning_rate": 5.053125690336127e-06,
+      "loss": 0.5654,
+      "step": 8033
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.758841269158062,
+      "learning_rate": 5.0520941612309425e-06,
+      "loss": 0.5102,
+      "step": 8034
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.0174381773721153,
+      "learning_rate": 5.051062629908276e-06,
+      "loss": 0.4772,
+      "step": 8035
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7727389710012816,
+      "learning_rate": 5.050031096412036e-06,
+      "loss": 0.4727,
+      "step": 8036
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6869542875623863,
+      "learning_rate": 5.048999560786132e-06,
+      "loss": 0.4683,
+      "step": 8037
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7761629624732608,
+      "learning_rate": 5.047968023074474e-06,
+      "loss": 0.4871,
+      "step": 8038
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6529765213071616,
+      "learning_rate": 5.046936483320969e-06,
+      "loss": 0.4833,
+      "step": 8039
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8153465847016584,
+      "learning_rate": 5.045904941569529e-06,
+      "loss": 0.5018,
+      "step": 8040
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5779988652875337,
+      "learning_rate": 5.044873397864063e-06,
+      "loss": 0.4747,
+      "step": 8041
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6758674169813761,
+      "learning_rate": 5.0438418522484785e-06,
+      "loss": 0.4542,
+      "step": 8042
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.9415339132924784,
+      "learning_rate": 5.042810304766688e-06,
+      "loss": 0.4939,
+      "step": 8043
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7408326508652936,
+      "learning_rate": 5.0417787554625984e-06,
+      "loss": 0.5083,
+      "step": 8044
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5427918811238637,
+      "learning_rate": 5.040747204380121e-06,
+      "loss": 0.4432,
+      "step": 8045
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6481608688800259,
+      "learning_rate": 5.0397156515631654e-06,
+      "loss": 0.4299,
+      "step": 8046
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5796860423287302,
+      "learning_rate": 5.038684097055641e-06,
+      "loss": 0.4965,
+      "step": 8047
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7174915973773783,
+      "learning_rate": 5.0376525409014585e-06,
+      "loss": 0.4193,
+      "step": 8048
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6954211386399821,
+      "learning_rate": 5.036620983144528e-06,
+      "loss": 0.4754,
+      "step": 8049
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7084335458577081,
+      "learning_rate": 5.03558942382876e-06,
+      "loss": 0.4233,
+      "step": 8050
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6916893243234785,
+      "learning_rate": 5.0345578629980605e-06,
+      "loss": 0.4413,
+      "step": 8051
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.0630844949586986,
+      "learning_rate": 5.033526300696346e-06,
+      "loss": 0.526,
+      "step": 8052
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.573586967354869,
+      "learning_rate": 5.032494736967525e-06,
+      "loss": 0.5498,
+      "step": 8053
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7079517205761564,
+      "learning_rate": 5.031463171855505e-06,
+      "loss": 0.4786,
+      "step": 8054
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6865691109249946,
+      "learning_rate": 5.030431605404199e-06,
+      "loss": 0.4786,
+      "step": 8055
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6232575626183916,
+      "learning_rate": 5.029400037657517e-06,
+      "loss": 0.5006,
+      "step": 8056
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8902358097940759,
+      "learning_rate": 5.028368468659368e-06,
+      "loss": 0.5555,
+      "step": 8057
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6793337767837564,
+      "learning_rate": 5.027336898453665e-06,
+      "loss": 0.5342,
+      "step": 8058
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5939482657068546,
+      "learning_rate": 5.026305327084318e-06,
+      "loss": 0.5256,
+      "step": 8059
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5916607362227726,
+      "learning_rate": 5.025273754595237e-06,
+      "loss": 0.4782,
+      "step": 8060
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.69822646169608,
+      "learning_rate": 5.024242181030332e-06,
+      "loss": 0.5008,
+      "step": 8061
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.752168127653501,
+      "learning_rate": 5.023210606433516e-06,
+      "loss": 0.5192,
+      "step": 8062
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9104293343091803,
+      "learning_rate": 5.022179030848698e-06,
+      "loss": 0.4992,
+      "step": 8063
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6680361321533085,
+      "learning_rate": 5.02114745431979e-06,
+      "loss": 0.4933,
+      "step": 8064
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5704580273500565,
+      "learning_rate": 5.020115876890702e-06,
+      "loss": 0.4279,
+      "step": 8065
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5712671736428463,
+      "learning_rate": 5.019084298605346e-06,
+      "loss": 0.448,
+      "step": 8066
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.818470240512177,
+      "learning_rate": 5.018052719507632e-06,
+      "loss": 0.5086,
+      "step": 8067
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6867642127681672,
+      "learning_rate": 5.0170211396414726e-06,
+      "loss": 0.5067,
+      "step": 8068
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8538337240545483,
+      "learning_rate": 5.015989559050777e-06,
+      "loss": 0.5239,
+      "step": 8069
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7331282310856594,
+      "learning_rate": 5.014957977779455e-06,
+      "loss": 0.5354,
+      "step": 8070
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8696451510505061,
+      "learning_rate": 5.013926395871421e-06,
+      "loss": 0.5801,
+      "step": 8071
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7634593629744055,
+      "learning_rate": 5.012894813370586e-06,
+      "loss": 0.4981,
+      "step": 8072
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5974869664160931,
+      "learning_rate": 5.0118632303208595e-06,
+      "loss": 0.4292,
+      "step": 8073
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7414298773320958,
+      "learning_rate": 5.0108316467661525e-06,
+      "loss": 0.504,
+      "step": 8074
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7498987872156508,
+      "learning_rate": 5.0098000627503775e-06,
+      "loss": 0.4294,
+      "step": 8075
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9569737905633081,
+      "learning_rate": 5.008768478317443e-06,
+      "loss": 0.4443,
+      "step": 8076
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6765939635966391,
+      "learning_rate": 5.007736893511265e-06,
+      "loss": 0.5097,
+      "step": 8077
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6107574568996386,
+      "learning_rate": 5.0067053083757515e-06,
+      "loss": 0.4749,
+      "step": 8078
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.54428483955781,
+      "learning_rate": 5.005673722954815e-06,
+      "loss": 0.4284,
+      "step": 8079
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.577817880241673,
+      "learning_rate": 5.004642137292365e-06,
+      "loss": 0.5118,
+      "step": 8080
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7605463303602441,
+      "learning_rate": 5.003610551432315e-06,
+      "loss": 0.4798,
+      "step": 8081
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6804233083766142,
+      "learning_rate": 5.002578965418575e-06,
+      "loss": 0.536,
+      "step": 8082
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7684880567027497,
+      "learning_rate": 5.001547379295057e-06,
+      "loss": 0.554,
+      "step": 8083
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6047929180025671,
+      "learning_rate": 5.000515793105671e-06,
+      "loss": 0.4575,
+      "step": 8084
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.619301551951031,
+      "learning_rate": 4.99948420689433e-06,
+      "loss": 0.4825,
+      "step": 8085
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7209483154140803,
+      "learning_rate": 4.998452620704944e-06,
+      "loss": 0.5081,
+      "step": 8086
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7318865073274593,
+      "learning_rate": 4.997421034581427e-06,
+      "loss": 0.4698,
+      "step": 8087
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7233770727932185,
+      "learning_rate": 4.9963894485676865e-06,
+      "loss": 0.5,
+      "step": 8088
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7438788769480187,
+      "learning_rate": 4.995357862707636e-06,
+      "loss": 0.4994,
+      "step": 8089
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6377103809597473,
+      "learning_rate": 4.994326277045188e-06,
+      "loss": 0.4386,
+      "step": 8090
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6318927812348238,
+      "learning_rate": 4.993294691624249e-06,
+      "loss": 0.5101,
+      "step": 8091
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8017299554073789,
+      "learning_rate": 4.992263106488736e-06,
+      "loss": 0.5517,
+      "step": 8092
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7065999620554689,
+      "learning_rate": 4.991231521682557e-06,
+      "loss": 0.5332,
+      "step": 8093
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6921646859919652,
+      "learning_rate": 4.990199937249624e-06,
+      "loss": 0.5236,
+      "step": 8094
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6314241567649539,
+      "learning_rate": 4.989168353233849e-06,
+      "loss": 0.5216,
+      "step": 8095
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7870602537592161,
+      "learning_rate": 4.988136769679143e-06,
+      "loss": 0.5748,
+      "step": 8096
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8868492606936104,
+      "learning_rate": 4.987105186629416e-06,
+      "loss": 0.5435,
+      "step": 8097
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5882130175175727,
+      "learning_rate": 4.98607360412858e-06,
+      "loss": 0.4647,
+      "step": 8098
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6667964928118641,
+      "learning_rate": 4.985042022220546e-06,
+      "loss": 0.5304,
+      "step": 8099
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7894263840401056,
+      "learning_rate": 4.9840104409492264e-06,
+      "loss": 0.5129,
+      "step": 8100
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6598057151425882,
+      "learning_rate": 4.982978860358531e-06,
+      "loss": 0.5547,
+      "step": 8101
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7475065550265066,
+      "learning_rate": 4.98194728049237e-06,
+      "loss": 0.4956,
+      "step": 8102
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6108297537146267,
+      "learning_rate": 4.9809157013946565e-06,
+      "loss": 0.5047,
+      "step": 8103
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8178384861045024,
+      "learning_rate": 4.979884123109298e-06,
+      "loss": 0.5725,
+      "step": 8104
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5955650242377624,
+      "learning_rate": 4.978852545680211e-06,
+      "loss": 0.4829,
+      "step": 8105
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5324535820881771,
+      "learning_rate": 4.977820969151302e-06,
+      "loss": 0.4561,
+      "step": 8106
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5801369673208406,
+      "learning_rate": 4.976789393566485e-06,
+      "loss": 0.4415,
+      "step": 8107
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5926204253813472,
+      "learning_rate": 4.975757818969669e-06,
+      "loss": 0.4245,
+      "step": 8108
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7496313150571022,
+      "learning_rate": 4.974726245404764e-06,
+      "loss": 0.4558,
+      "step": 8109
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6052156373739525,
+      "learning_rate": 4.973694672915684e-06,
+      "loss": 0.4904,
+      "step": 8110
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.668297626818539,
+      "learning_rate": 4.972663101546337e-06,
+      "loss": 0.5069,
+      "step": 8111
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7076046746714018,
+      "learning_rate": 4.9716315313406336e-06,
+      "loss": 0.5435,
+      "step": 8112
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.783893601648489,
+      "learning_rate": 4.970599962342486e-06,
+      "loss": 0.537,
+      "step": 8113
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6254790843764756,
+      "learning_rate": 4.969568394595803e-06,
+      "loss": 0.4416,
+      "step": 8114
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.0921393311376757,
+      "learning_rate": 4.968536828144497e-06,
+      "loss": 0.5504,
+      "step": 8115
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7745381813738811,
+      "learning_rate": 4.967505263032476e-06,
+      "loss": 0.5733,
+      "step": 8116
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7768241016381884,
+      "learning_rate": 4.966473699303654e-06,
+      "loss": 0.4973,
+      "step": 8117
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5180697857474678,
+      "learning_rate": 4.965442137001939e-06,
+      "loss": 0.4182,
+      "step": 8118
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6311950676772196,
+      "learning_rate": 4.964410576171243e-06,
+      "loss": 0.4924,
+      "step": 8119
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7820784613143212,
+      "learning_rate": 4.9633790168554735e-06,
+      "loss": 0.4804,
+      "step": 8120
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7414341015896759,
+      "learning_rate": 4.962347459098542e-06,
+      "loss": 0.5369,
+      "step": 8121
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8201712140698587,
+      "learning_rate": 4.96131590294436e-06,
+      "loss": 0.4781,
+      "step": 8122
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6621278595704478,
+      "learning_rate": 4.960284348436837e-06,
+      "loss": 0.478,
+      "step": 8123
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7318612303570251,
+      "learning_rate": 4.959252795619881e-06,
+      "loss": 0.4795,
+      "step": 8124
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6843648599542215,
+      "learning_rate": 4.958221244537404e-06,
+      "loss": 0.4949,
+      "step": 8125
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8638883665995928,
+      "learning_rate": 4.9571896952333145e-06,
+      "loss": 0.5158,
+      "step": 8126
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8735920736913573,
+      "learning_rate": 4.956158147751523e-06,
+      "loss": 0.4905,
+      "step": 8127
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5738007049937406,
+      "learning_rate": 4.955126602135938e-06,
+      "loss": 0.4868,
+      "step": 8128
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6219592933355601,
+      "learning_rate": 4.954095058430471e-06,
+      "loss": 0.4739,
+      "step": 8129
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7957876962280005,
+      "learning_rate": 4.953063516679031e-06,
+      "loss": 0.5083,
+      "step": 8130
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6263593027929184,
+      "learning_rate": 4.952031976925528e-06,
+      "loss": 0.469,
+      "step": 8131
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6326247401095945,
+      "learning_rate": 4.951000439213869e-06,
+      "loss": 0.4568,
+      "step": 8132
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5686671787418717,
+      "learning_rate": 4.949968903587966e-06,
+      "loss": 0.4107,
+      "step": 8133
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6770448684844186,
+      "learning_rate": 4.948937370091726e-06,
+      "loss": 0.4586,
+      "step": 8134
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6198756006943074,
+      "learning_rate": 4.947905838769059e-06,
+      "loss": 0.4489,
+      "step": 8135
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.588322239979068,
+      "learning_rate": 4.946874309663875e-06,
+      "loss": 0.4786,
+      "step": 8136
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7281434699718707,
+      "learning_rate": 4.945842782820081e-06,
+      "loss": 0.4367,
+      "step": 8137
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6656070348463957,
+      "learning_rate": 4.944811258281586e-06,
+      "loss": 0.4938,
+      "step": 8138
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.769071296361864,
+      "learning_rate": 4.9437797360923005e-06,
+      "loss": 0.5517,
+      "step": 8139
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.679566649812272,
+      "learning_rate": 4.942748216296132e-06,
+      "loss": 0.483,
+      "step": 8140
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7801902138019788,
+      "learning_rate": 4.941716698936987e-06,
+      "loss": 0.5542,
+      "step": 8141
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.588084511685038,
+      "learning_rate": 4.940685184058778e-06,
+      "loss": 0.4881,
+      "step": 8142
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7338863472434032,
+      "learning_rate": 4.93965367170541e-06,
+      "loss": 0.4935,
+      "step": 8143
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5640269128996757,
+      "learning_rate": 4.938622161920793e-06,
+      "loss": 0.4243,
+      "step": 8144
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6597540894818068,
+      "learning_rate": 4.937590654748835e-06,
+      "loss": 0.4653,
+      "step": 8145
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6558563852646573,
+      "learning_rate": 4.936559150233443e-06,
+      "loss": 0.4711,
+      "step": 8146
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6021679819291212,
+      "learning_rate": 4.935527648418524e-06,
+      "loss": 0.4561,
+      "step": 8147
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.774985474628311,
+      "learning_rate": 4.9344961493479885e-06,
+      "loss": 0.4616,
+      "step": 8148
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.790357340425759,
+      "learning_rate": 4.9334646530657415e-06,
+      "loss": 0.5451,
+      "step": 8149
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6770652021077532,
+      "learning_rate": 4.932433159615693e-06,
+      "loss": 0.4564,
+      "step": 8150
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6783212701251177,
+      "learning_rate": 4.931401669041748e-06,
+      "loss": 0.4789,
+      "step": 8151
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.796682217235123,
+      "learning_rate": 4.9303701813878144e-06,
+      "loss": 0.5437,
+      "step": 8152
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.0071332259830925,
+      "learning_rate": 4.9293386966977994e-06,
+      "loss": 0.5755,
+      "step": 8153
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.5310500434845555,
+      "learning_rate": 4.928307215015611e-06,
+      "loss": 0.5354,
+      "step": 8154
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6400968540349614,
+      "learning_rate": 4.9272757363851555e-06,
+      "loss": 0.4971,
+      "step": 8155
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7638101579938189,
+      "learning_rate": 4.92624426085034e-06,
+      "loss": 0.5345,
+      "step": 8156
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6978492382260936,
+      "learning_rate": 4.92521278845507e-06,
+      "loss": 0.4906,
+      "step": 8157
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7935373841448372,
+      "learning_rate": 4.924181319243253e-06,
+      "loss": 0.5321,
+      "step": 8158
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8144939574626947,
+      "learning_rate": 4.923149853258795e-06,
+      "loss": 0.5151,
+      "step": 8159
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5847324438157572,
+      "learning_rate": 4.922118390545602e-06,
+      "loss": 0.4286,
+      "step": 8160
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5837197646959561,
+      "learning_rate": 4.92108693114758e-06,
+      "loss": 0.4735,
+      "step": 8161
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6576688912176004,
+      "learning_rate": 4.9200554751086354e-06,
+      "loss": 0.472,
+      "step": 8162
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7806261387489312,
+      "learning_rate": 4.919024022472674e-06,
+      "loss": 0.5474,
+      "step": 8163
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7440678771464696,
+      "learning_rate": 4.9179925732836e-06,
+      "loss": 0.4915,
+      "step": 8164
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6270807035856898,
+      "learning_rate": 4.916961127585322e-06,
+      "loss": 0.4168,
+      "step": 8165
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6062933713288158,
+      "learning_rate": 4.91592968542174e-06,
+      "loss": 0.4253,
+      "step": 8166
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8093407222557867,
+      "learning_rate": 4.914898246836764e-06,
+      "loss": 0.5136,
+      "step": 8167
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6658720365941423,
+      "learning_rate": 4.9138668118742994e-06,
+      "loss": 0.5368,
+      "step": 8168
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8033865420421944,
+      "learning_rate": 4.912835380578249e-06,
+      "loss": 0.5311,
+      "step": 8169
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7002685475646482,
+      "learning_rate": 4.911803952992518e-06,
+      "loss": 0.496,
+      "step": 8170
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6824569246723416,
+      "learning_rate": 4.910772529161009e-06,
+      "loss": 0.4896,
+      "step": 8171
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6926253996415783,
+      "learning_rate": 4.909741109127631e-06,
+      "loss": 0.534,
+      "step": 8172
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.101069266180054,
+      "learning_rate": 4.908709692936284e-06,
+      "loss": 0.4662,
+      "step": 8173
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7384209669025243,
+      "learning_rate": 4.907678280630874e-06,
+      "loss": 0.491,
+      "step": 8174
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6977103742536973,
+      "learning_rate": 4.906646872255305e-06,
+      "loss": 0.5196,
+      "step": 8175
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.628473299941772,
+      "learning_rate": 4.90561546785348e-06,
+      "loss": 0.5175,
+      "step": 8176
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8071292100474738,
+      "learning_rate": 4.904584067469303e-06,
+      "loss": 0.5382,
+      "step": 8177
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7617304133126185,
+      "learning_rate": 4.903552671146675e-06,
+      "loss": 0.581,
+      "step": 8178
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7305998341939557,
+      "learning_rate": 4.902521278929504e-06,
+      "loss": 0.4861,
+      "step": 8179
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5931602093389495,
+      "learning_rate": 4.90148989086169e-06,
+      "loss": 0.4389,
+      "step": 8180
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8133656696254687,
+      "learning_rate": 4.900458506987137e-06,
+      "loss": 0.5124,
+      "step": 8181
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7196123470267825,
+      "learning_rate": 4.899427127349747e-06,
+      "loss": 0.4968,
+      "step": 8182
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6756696081050716,
+      "learning_rate": 4.898395751993423e-06,
+      "loss": 0.4096,
+      "step": 8183
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6306960210407471,
+      "learning_rate": 4.897364380962068e-06,
+      "loss": 0.4724,
+      "step": 8184
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6040945195355162,
+      "learning_rate": 4.8963330142995826e-06,
+      "loss": 0.4662,
+      "step": 8185
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7188606257204347,
+      "learning_rate": 4.895301652049869e-06,
+      "loss": 0.5029,
+      "step": 8186
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6713362921139758,
+      "learning_rate": 4.8942702942568305e-06,
+      "loss": 0.456,
+      "step": 8187
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7191827252342243,
+      "learning_rate": 4.893238940964367e-06,
+      "loss": 0.5407,
+      "step": 8188
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7805356564484168,
+      "learning_rate": 4.8922075922163804e-06,
+      "loss": 0.5193,
+      "step": 8189
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.822079346452177,
+      "learning_rate": 4.891176248056771e-06,
+      "loss": 0.5152,
+      "step": 8190
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7058846562952281,
+      "learning_rate": 4.890144908529442e-06,
+      "loss": 0.5226,
+      "step": 8191
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.64522121209394,
+      "learning_rate": 4.889113573678294e-06,
+      "loss": 0.5133,
+      "step": 8192
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7002708813522056,
+      "learning_rate": 4.888082243547226e-06,
+      "loss": 0.5312,
+      "step": 8193
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.764951850368504,
+      "learning_rate": 4.88705091818014e-06,
+      "loss": 0.5119,
+      "step": 8194
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.76976036414379,
+      "learning_rate": 4.8860195976209354e-06,
+      "loss": 0.5069,
+      "step": 8195
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8286293076818172,
+      "learning_rate": 4.884988281913512e-06,
+      "loss": 0.5364,
+      "step": 8196
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7572316449150073,
+      "learning_rate": 4.88395697110177e-06,
+      "loss": 0.4912,
+      "step": 8197
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7723155499331286,
+      "learning_rate": 4.88292566522961e-06,
+      "loss": 0.4772,
+      "step": 8198
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6078057153870111,
+      "learning_rate": 4.881894364340929e-06,
+      "loss": 0.517,
+      "step": 8199
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7053595877798743,
+      "learning_rate": 4.880863068479628e-06,
+      "loss": 0.5052,
+      "step": 8200
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6593743158834144,
+      "learning_rate": 4.879831777689606e-06,
+      "loss": 0.4666,
+      "step": 8201
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6361065123969063,
+      "learning_rate": 4.87880049201476e-06,
+      "loss": 0.4491,
+      "step": 8202
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5716349027885537,
+      "learning_rate": 4.877769211498989e-06,
+      "loss": 0.4283,
+      "step": 8203
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5720540039427114,
+      "learning_rate": 4.876737936186193e-06,
+      "loss": 0.4775,
+      "step": 8204
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6165947747468568,
+      "learning_rate": 4.875706666120269e-06,
+      "loss": 0.4801,
+      "step": 8205
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.792961072769179,
+      "learning_rate": 4.874675401345116e-06,
+      "loss": 0.5268,
+      "step": 8206
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.815287152901661,
+      "learning_rate": 4.8736441419046305e-06,
+      "loss": 0.4854,
+      "step": 8207
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.9903397150894573,
+      "learning_rate": 4.87261288784271e-06,
+      "loss": 0.5335,
+      "step": 8208
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7595958288211314,
+      "learning_rate": 4.871581639203251e-06,
+      "loss": 0.5459,
+      "step": 8209
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6256464266848701,
+      "learning_rate": 4.8705503960301515e-06,
+      "loss": 0.4494,
+      "step": 8210
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6687912366964109,
+      "learning_rate": 4.869519158367308e-06,
+      "loss": 0.4776,
+      "step": 8211
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.801744174634166,
+      "learning_rate": 4.8684879262586175e-06,
+      "loss": 0.5669,
+      "step": 8212
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7645578124566696,
+      "learning_rate": 4.867456699747975e-06,
+      "loss": 0.5607,
+      "step": 8213
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7216416974605512,
+      "learning_rate": 4.866425478879279e-06,
+      "loss": 0.5119,
+      "step": 8214
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7459689271359239,
+      "learning_rate": 4.86539426369642e-06,
+      "loss": 0.4735,
+      "step": 8215
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6405112781348862,
+      "learning_rate": 4.8643630542433005e-06,
+      "loss": 0.5296,
+      "step": 8216
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8086311427381001,
+      "learning_rate": 4.863331850563811e-06,
+      "loss": 0.5644,
+      "step": 8217
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.0809850199777922,
+      "learning_rate": 4.8623006527018475e-06,
+      "loss": 0.5506,
+      "step": 8218
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8355964877091129,
+      "learning_rate": 4.861269460701306e-06,
+      "loss": 0.4848,
+      "step": 8219
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6083604806636115,
+      "learning_rate": 4.86023827460608e-06,
+      "loss": 0.4785,
+      "step": 8220
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8007060967062233,
+      "learning_rate": 4.859207094460065e-06,
+      "loss": 0.5167,
+      "step": 8221
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8139772792869403,
+      "learning_rate": 4.858175920307153e-06,
+      "loss": 0.5315,
+      "step": 8222
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6990414904321893,
+      "learning_rate": 4.857144752191238e-06,
+      "loss": 0.5069,
+      "step": 8223
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7490374445137454,
+      "learning_rate": 4.856113590156216e-06,
+      "loss": 0.5491,
+      "step": 8224
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7928538358900609,
+      "learning_rate": 4.855082434245978e-06,
+      "loss": 0.5334,
+      "step": 8225
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.664875956858229,
+      "learning_rate": 4.854051284504418e-06,
+      "loss": 0.5,
+      "step": 8226
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8205903752505009,
+      "learning_rate": 4.8530201409754285e-06,
+      "loss": 0.5423,
+      "step": 8227
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8122183398227888,
+      "learning_rate": 4.8519890037029e-06,
+      "loss": 0.5103,
+      "step": 8228
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6594882528654233,
+      "learning_rate": 4.850957872730728e-06,
+      "loss": 0.4827,
+      "step": 8229
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6273417028965672,
+      "learning_rate": 4.849926748102803e-06,
+      "loss": 0.4645,
+      "step": 8230
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6912005844276828,
+      "learning_rate": 4.848895629863018e-06,
+      "loss": 0.4743,
+      "step": 8231
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7023085892697838,
+      "learning_rate": 4.847864518055261e-06,
+      "loss": 0.4735,
+      "step": 8232
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5654344139350759,
+      "learning_rate": 4.8468334127234275e-06,
+      "loss": 0.4346,
+      "step": 8233
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6681639496497829,
+      "learning_rate": 4.845802313911405e-06,
+      "loss": 0.4418,
+      "step": 8234
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6608505034354452,
+      "learning_rate": 4.844771221663086e-06,
+      "loss": 0.4885,
+      "step": 8235
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6250645790722317,
+      "learning_rate": 4.843740136022359e-06,
+      "loss": 0.5025,
+      "step": 8236
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7429447805268162,
+      "learning_rate": 4.842709057033116e-06,
+      "loss": 0.4715,
+      "step": 8237
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5633102581255295,
+      "learning_rate": 4.841677984739245e-06,
+      "loss": 0.4234,
+      "step": 8238
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5980258866552804,
+      "learning_rate": 4.8406469191846374e-06,
+      "loss": 0.4852,
+      "step": 8239
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7968220275413964,
+      "learning_rate": 4.839615860413178e-06,
+      "loss": 0.5105,
+      "step": 8240
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5617390826528584,
+      "learning_rate": 4.838584808468761e-06,
+      "loss": 0.4465,
+      "step": 8241
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6124345282934126,
+      "learning_rate": 4.837553763395274e-06,
+      "loss": 0.4664,
+      "step": 8242
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.529315244232884,
+      "learning_rate": 4.836522725236604e-06,
+      "loss": 0.4149,
+      "step": 8243
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.778035947403778,
+      "learning_rate": 4.835491694036638e-06,
+      "loss": 0.4814,
+      "step": 8244
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6791867705029339,
+      "learning_rate": 4.834460669839266e-06,
+      "loss": 0.5016,
+      "step": 8245
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5673865733446531,
+      "learning_rate": 4.833429652688374e-06,
+      "loss": 0.4406,
+      "step": 8246
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.53186758935031,
+      "learning_rate": 4.832398642627849e-06,
+      "loss": 0.4186,
+      "step": 8247
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6744997106497785,
+      "learning_rate": 4.831367639701579e-06,
+      "loss": 0.4676,
+      "step": 8248
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7006821578003772,
+      "learning_rate": 4.830336643953449e-06,
+      "loss": 0.5783,
+      "step": 8249
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7494961788681345,
+      "learning_rate": 4.829305655427346e-06,
+      "loss": 0.4503,
+      "step": 8250
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7960997699671731,
+      "learning_rate": 4.828274674167156e-06,
+      "loss": 0.5512,
+      "step": 8251
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7252706700449726,
+      "learning_rate": 4.827243700216762e-06,
+      "loss": 0.5438,
+      "step": 8252
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6124893123667524,
+      "learning_rate": 4.826212733620054e-06,
+      "loss": 0.4649,
+      "step": 8253
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6201293585842711,
+      "learning_rate": 4.825181774420915e-06,
+      "loss": 0.4471,
+      "step": 8254
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6010911318106227,
+      "learning_rate": 4.8241508226632285e-06,
+      "loss": 0.4478,
+      "step": 8255
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6629589461654971,
+      "learning_rate": 4.82311987839088e-06,
+      "loss": 0.468,
+      "step": 8256
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7122458328776413,
+      "learning_rate": 4.822088941647753e-06,
+      "loss": 0.4981,
+      "step": 8257
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7225105454312586,
+      "learning_rate": 4.821058012477731e-06,
+      "loss": 0.4917,
+      "step": 8258
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6356762569082717,
+      "learning_rate": 4.820027090924698e-06,
+      "loss": 0.5199,
+      "step": 8259
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7390531461066636,
+      "learning_rate": 4.818996177032536e-06,
+      "loss": 0.4906,
+      "step": 8260
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6715414809108625,
+      "learning_rate": 4.817965270845129e-06,
+      "loss": 0.4907,
+      "step": 8261
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7872491401963878,
+      "learning_rate": 4.8169343724063574e-06,
+      "loss": 0.521,
+      "step": 8262
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8312153903219698,
+      "learning_rate": 4.8159034817601055e-06,
+      "loss": 0.5818,
+      "step": 8263
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7546787280603171,
+      "learning_rate": 4.814872598950255e-06,
+      "loss": 0.5496,
+      "step": 8264
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7716455672093985,
+      "learning_rate": 4.813841724020684e-06,
+      "loss": 0.5266,
+      "step": 8265
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7240966792837004,
+      "learning_rate": 4.812810857015278e-06,
+      "loss": 0.5492,
+      "step": 8266
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6694567122072812,
+      "learning_rate": 4.811779997977914e-06,
+      "loss": 0.4808,
+      "step": 8267
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8307680593191434,
+      "learning_rate": 4.8107491469524756e-06,
+      "loss": 0.5309,
+      "step": 8268
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6751536402282458,
+      "learning_rate": 4.809718303982841e-06,
+      "loss": 0.4805,
+      "step": 8269
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6384524355478715,
+      "learning_rate": 4.8086874691128896e-06,
+      "loss": 0.4779,
+      "step": 8270
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8092827043226495,
+      "learning_rate": 4.807656642386501e-06,
+      "loss": 0.5593,
+      "step": 8271
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6841599902671923,
+      "learning_rate": 4.806625823847555e-06,
+      "loss": 0.4846,
+      "step": 8272
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6388523843509164,
+      "learning_rate": 4.80559501353993e-06,
+      "loss": 0.4693,
+      "step": 8273
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.686568112462965,
+      "learning_rate": 4.804564211507504e-06,
+      "loss": 0.4866,
+      "step": 8274
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7377607432941956,
+      "learning_rate": 4.803533417794155e-06,
+      "loss": 0.5379,
+      "step": 8275
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6125103250502069,
+      "learning_rate": 4.80250263244376e-06,
+      "loss": 0.45,
+      "step": 8276
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.612603928040167,
+      "learning_rate": 4.8014718555001964e-06,
+      "loss": 0.5093,
+      "step": 8277
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6598029494031642,
+      "learning_rate": 4.800441087007342e-06,
+      "loss": 0.5087,
+      "step": 8278
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6311228028854338,
+      "learning_rate": 4.7994103270090735e-06,
+      "loss": 0.5057,
+      "step": 8279
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7268557525529523,
+      "learning_rate": 4.798379575549266e-06,
+      "loss": 0.5394,
+      "step": 8280
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5701984099904688,
+      "learning_rate": 4.797348832671796e-06,
+      "loss": 0.4342,
+      "step": 8281
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6421093650108104,
+      "learning_rate": 4.796318098420538e-06,
+      "loss": 0.4723,
+      "step": 8282
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8177789621404639,
+      "learning_rate": 4.795287372839368e-06,
+      "loss": 0.5494,
+      "step": 8283
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.9362342834889982,
+      "learning_rate": 4.794256655972161e-06,
+      "loss": 0.5064,
+      "step": 8284
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7045524933158974,
+      "learning_rate": 4.79322594786279e-06,
+      "loss": 0.509,
+      "step": 8285
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8361017931108989,
+      "learning_rate": 4.7921952485551295e-06,
+      "loss": 0.5002,
+      "step": 8286
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6032002726730082,
+      "learning_rate": 4.791164558093054e-06,
+      "loss": 0.4493,
+      "step": 8287
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6165107148119306,
+      "learning_rate": 4.790133876520435e-06,
+      "loss": 0.4552,
+      "step": 8288
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.721207995116974,
+      "learning_rate": 4.789103203881147e-06,
+      "loss": 0.5762,
+      "step": 8289
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7628121153743485,
+      "learning_rate": 4.7880725402190595e-06,
+      "loss": 0.5021,
+      "step": 8290
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8310262869235524,
+      "learning_rate": 4.787041885578048e-06,
+      "loss": 0.5734,
+      "step": 8291
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.774089249774509,
+      "learning_rate": 4.7860112400019834e-06,
+      "loss": 0.5021,
+      "step": 8292
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6050785921716553,
+      "learning_rate": 4.784980603534737e-06,
+      "loss": 0.4829,
+      "step": 8293
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6850084566210694,
+      "learning_rate": 4.783949976220179e-06,
+      "loss": 0.5018,
+      "step": 8294
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.0573215637059092,
+      "learning_rate": 4.782919358102179e-06,
+      "loss": 0.4795,
+      "step": 8295
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6547241669697361,
+      "learning_rate": 4.78188874922461e-06,
+      "loss": 0.4866,
+      "step": 8296
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6478676320357795,
+      "learning_rate": 4.7808581496313385e-06,
+      "loss": 0.4525,
+      "step": 8297
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6843452654307315,
+      "learning_rate": 4.779827559366236e-06,
+      "loss": 0.4987,
+      "step": 8298
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.749033531854335,
+      "learning_rate": 4.778796978473171e-06,
+      "loss": 0.4797,
+      "step": 8299
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8176461849848516,
+      "learning_rate": 4.777766406996011e-06,
+      "loss": 0.5129,
+      "step": 8300
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5705261620536343,
+      "learning_rate": 4.776735844978626e-06,
+      "loss": 0.4314,
+      "step": 8301
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6700612230120924,
+      "learning_rate": 4.77570529246488e-06,
+      "loss": 0.4941,
+      "step": 8302
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6228444338642064,
+      "learning_rate": 4.774674749498645e-06,
+      "loss": 0.5081,
+      "step": 8303
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7172664407984548,
+      "learning_rate": 4.773644216123785e-06,
+      "loss": 0.5206,
+      "step": 8304
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7464573812938363,
+      "learning_rate": 4.772613692384168e-06,
+      "loss": 0.5402,
+      "step": 8305
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7071676505134432,
+      "learning_rate": 4.77158317832366e-06,
+      "loss": 0.5094,
+      "step": 8306
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5489632515712463,
+      "learning_rate": 4.770552673986125e-06,
+      "loss": 0.4181,
+      "step": 8307
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7151312126475369,
+      "learning_rate": 4.7695221794154315e-06,
+      "loss": 0.5255,
+      "step": 8308
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8168589284014309,
+      "learning_rate": 4.768491694655441e-06,
+      "loss": 0.5374,
+      "step": 8309
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7398942962946595,
+      "learning_rate": 4.7674612197500194e-06,
+      "loss": 0.5418,
+      "step": 8310
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6906132880351743,
+      "learning_rate": 4.76643075474303e-06,
+      "loss": 0.5171,
+      "step": 8311
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7496395408570655,
+      "learning_rate": 4.7654002996783375e-06,
+      "loss": 0.4947,
+      "step": 8312
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6958341591522413,
+      "learning_rate": 4.764369854599805e-06,
+      "loss": 0.5191,
+      "step": 8313
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7685756429343458,
+      "learning_rate": 4.763339419551292e-06,
+      "loss": 0.5214,
+      "step": 8314
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6372399625193196,
+      "learning_rate": 4.762308994576666e-06,
+      "loss": 0.4885,
+      "step": 8315
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6102657045220327,
+      "learning_rate": 4.7612785797197865e-06,
+      "loss": 0.4684,
+      "step": 8316
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6579304653084039,
+      "learning_rate": 4.760248175024515e-06,
+      "loss": 0.4952,
+      "step": 8317
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7343753218958654,
+      "learning_rate": 4.759217780534713e-06,
+      "loss": 0.4803,
+      "step": 8318
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7741027393340881,
+      "learning_rate": 4.758187396294241e-06,
+      "loss": 0.529,
+      "step": 8319
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7667035487416406,
+      "learning_rate": 4.7571570223469575e-06,
+      "loss": 0.5363,
+      "step": 8320
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6107481436895474,
+      "learning_rate": 4.756126658736725e-06,
+      "loss": 0.4528,
+      "step": 8321
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6737471703064847,
+      "learning_rate": 4.7550963055074e-06,
+      "loss": 0.4782,
+      "step": 8322
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8540866683262708,
+      "learning_rate": 4.754065962702843e-06,
+      "loss": 0.532,
+      "step": 8323
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.1902404904276669,
+      "learning_rate": 4.753035630366913e-06,
+      "loss": 0.5155,
+      "step": 8324
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5897317747309352,
+      "learning_rate": 4.752005308543466e-06,
+      "loss": 0.4303,
+      "step": 8325
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7012971327486279,
+      "learning_rate": 4.750974997276361e-06,
+      "loss": 0.4871,
+      "step": 8326
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7640488877322428,
+      "learning_rate": 4.749944696609453e-06,
+      "loss": 0.541,
+      "step": 8327
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7737930288919439,
+      "learning_rate": 4.748914406586602e-06,
+      "loss": 0.5053,
+      "step": 8328
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5911253142058805,
+      "learning_rate": 4.7478841272516616e-06,
+      "loss": 0.4478,
+      "step": 8329
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5795290210295466,
+      "learning_rate": 4.746853858648489e-06,
+      "loss": 0.4683,
+      "step": 8330
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6489249567019832,
+      "learning_rate": 4.745823600820939e-06,
+      "loss": 0.4687,
+      "step": 8331
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7705164626644723,
+      "learning_rate": 4.7447933538128634e-06,
+      "loss": 0.5312,
+      "step": 8332
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6608120498356257,
+      "learning_rate": 4.743763117668121e-06,
+      "loss": 0.5045,
+      "step": 8333
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7930373929415717,
+      "learning_rate": 4.742732892430565e-06,
+      "loss": 0.5384,
+      "step": 8334
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7543329925328313,
+      "learning_rate": 4.741702678144047e-06,
+      "loss": 0.5222,
+      "step": 8335
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5520664241237638,
+      "learning_rate": 4.74067247485242e-06,
+      "loss": 0.4485,
+      "step": 8336
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7122324134902953,
+      "learning_rate": 4.739642282599538e-06,
+      "loss": 0.4918,
+      "step": 8337
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.680478604032049,
+      "learning_rate": 4.7386121014292505e-06,
+      "loss": 0.5085,
+      "step": 8338
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5600107710486806,
+      "learning_rate": 4.737581931385411e-06,
+      "loss": 0.4584,
+      "step": 8339
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6896580366720484,
+      "learning_rate": 4.73655177251187e-06,
+      "loss": 0.4906,
+      "step": 8340
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.963977787932417,
+      "learning_rate": 4.73552162485248e-06,
+      "loss": 0.5213,
+      "step": 8341
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7043908635505579,
+      "learning_rate": 4.734491488451087e-06,
+      "loss": 0.4862,
+      "step": 8342
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6881430654845729,
+      "learning_rate": 4.733461363351544e-06,
+      "loss": 0.5056,
+      "step": 8343
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7424707899379075,
+      "learning_rate": 4.7324312495976994e-06,
+      "loss": 0.5498,
+      "step": 8344
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6172766748002575,
+      "learning_rate": 4.731401147233402e-06,
+      "loss": 0.471,
+      "step": 8345
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6407102122432393,
+      "learning_rate": 4.730371056302498e-06,
+      "loss": 0.4529,
+      "step": 8346
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6719462297994628,
+      "learning_rate": 4.7293409768488365e-06,
+      "loss": 0.4964,
+      "step": 8347
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6584934755649497,
+      "learning_rate": 4.728310908916266e-06,
+      "loss": 0.4518,
+      "step": 8348
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.662125686194116,
+      "learning_rate": 4.727280852548632e-06,
+      "loss": 0.4953,
+      "step": 8349
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6914782045059399,
+      "learning_rate": 4.726250807789779e-06,
+      "loss": 0.539,
+      "step": 8350
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6847308364637184,
+      "learning_rate": 4.725220774683555e-06,
+      "loss": 0.4303,
+      "step": 8351
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.598398400830368,
+      "learning_rate": 4.724190753273803e-06,
+      "loss": 0.4903,
+      "step": 8352
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7146368773370222,
+      "learning_rate": 4.723160743604371e-06,
+      "loss": 0.492,
+      "step": 8353
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.686039990221673,
+      "learning_rate": 4.7221307457191014e-06,
+      "loss": 0.4945,
+      "step": 8354
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6349715579352723,
+      "learning_rate": 4.721100759661838e-06,
+      "loss": 0.5377,
+      "step": 8355
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7781684646665405,
+      "learning_rate": 4.720070785476424e-06,
+      "loss": 0.5561,
+      "step": 8356
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8482692988535854,
+      "learning_rate": 4.719040823206702e-06,
+      "loss": 0.5792,
+      "step": 8357
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7842530178277606,
+      "learning_rate": 4.718010872896515e-06,
+      "loss": 0.5251,
+      "step": 8358
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7524235995930882,
+      "learning_rate": 4.716980934589703e-06,
+      "loss": 0.5123,
+      "step": 8359
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8094302828328926,
+      "learning_rate": 4.7159510083301095e-06,
+      "loss": 0.512,
+      "step": 8360
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5494737537226451,
+      "learning_rate": 4.714921094161573e-06,
+      "loss": 0.4264,
+      "step": 8361
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7242920027653952,
+      "learning_rate": 4.713891192127935e-06,
+      "loss": 0.538,
+      "step": 8362
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.702317673340286,
+      "learning_rate": 4.712861302273034e-06,
+      "loss": 0.4911,
+      "step": 8363
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6389843628300853,
+      "learning_rate": 4.7118314246407084e-06,
+      "loss": 0.4368,
+      "step": 8364
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6049971019201188,
+      "learning_rate": 4.7108015592748005e-06,
+      "loss": 0.4483,
+      "step": 8365
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6903272201243086,
+      "learning_rate": 4.709771706219145e-06,
+      "loss": 0.4091,
+      "step": 8366
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7160418672888511,
+      "learning_rate": 4.708741865517581e-06,
+      "loss": 0.5074,
+      "step": 8367
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6924792251940831,
+      "learning_rate": 4.7077120372139455e-06,
+      "loss": 0.4431,
+      "step": 8368
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6828736714010408,
+      "learning_rate": 4.706682221352074e-06,
+      "loss": 0.4541,
+      "step": 8369
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7528016280901391,
+      "learning_rate": 4.705652417975803e-06,
+      "loss": 0.5315,
+      "step": 8370
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6899681482329726,
+      "learning_rate": 4.704622627128969e-06,
+      "loss": 0.541,
+      "step": 8371
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8629331194423644,
+      "learning_rate": 4.703592848855405e-06,
+      "loss": 0.6034,
+      "step": 8372
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.765164791092945,
+      "learning_rate": 4.7025630831989465e-06,
+      "loss": 0.5053,
+      "step": 8373
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6877677701275337,
+      "learning_rate": 4.701533330203427e-06,
+      "loss": 0.5148,
+      "step": 8374
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7151067895192951,
+      "learning_rate": 4.70050358991268e-06,
+      "loss": 0.5062,
+      "step": 8375
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7062122950474459,
+      "learning_rate": 4.699473862370535e-06,
+      "loss": 0.4794,
+      "step": 8376
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5964550116294689,
+      "learning_rate": 4.698444147620831e-06,
+      "loss": 0.451,
+      "step": 8377
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5786820849146606,
+      "learning_rate": 4.697414445707395e-06,
+      "loss": 0.4311,
+      "step": 8378
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6020010219440008,
+      "learning_rate": 4.696384756674059e-06,
+      "loss": 0.519,
+      "step": 8379
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7530596084436874,
+      "learning_rate": 4.695355080564655e-06,
+      "loss": 0.523,
+      "step": 8380
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6512234287017661,
+      "learning_rate": 4.69432541742301e-06,
+      "loss": 0.4924,
+      "step": 8381
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7564569630147391,
+      "learning_rate": 4.6932957672929565e-06,
+      "loss": 0.5666,
+      "step": 8382
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.384880958266251,
+      "learning_rate": 4.692266130218322e-06,
+      "loss": 0.4818,
+      "step": 8383
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6994452017916899,
+      "learning_rate": 4.6912365062429334e-06,
+      "loss": 0.4573,
+      "step": 8384
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6832932871007706,
+      "learning_rate": 4.690206895410622e-06,
+      "loss": 0.4641,
+      "step": 8385
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8228968802336903,
+      "learning_rate": 4.689177297765212e-06,
+      "loss": 0.5673,
+      "step": 8386
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.754176995111916,
+      "learning_rate": 4.688147713350532e-06,
+      "loss": 0.5774,
+      "step": 8387
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8485384255773295,
+      "learning_rate": 4.687118142210407e-06,
+      "loss": 0.5713,
+      "step": 8388
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9218049433550063,
+      "learning_rate": 4.686088584388661e-06,
+      "loss": 0.5258,
+      "step": 8389
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7431226167112078,
+      "learning_rate": 4.685059039929123e-06,
+      "loss": 0.5334,
+      "step": 8390
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7642740832397854,
+      "learning_rate": 4.684029508875615e-06,
+      "loss": 0.5194,
+      "step": 8391
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6598599128505827,
+      "learning_rate": 4.682999991271961e-06,
+      "loss": 0.4935,
+      "step": 8392
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6323138682990631,
+      "learning_rate": 4.681970487161984e-06,
+      "loss": 0.496,
+      "step": 8393
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.4095994915849337,
+      "learning_rate": 4.680940996589506e-06,
+      "loss": 0.5157,
+      "step": 8394
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7512926439974584,
+      "learning_rate": 4.6799115195983515e-06,
+      "loss": 0.48,
+      "step": 8395
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7554170976018972,
+      "learning_rate": 4.678882056232341e-06,
+      "loss": 0.5167,
+      "step": 8396
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6039114237446409,
+      "learning_rate": 4.677852606535295e-06,
+      "loss": 0.4556,
+      "step": 8397
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5463591241004426,
+      "learning_rate": 4.6768231705510346e-06,
+      "loss": 0.4468,
+      "step": 8398
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6700800719790695,
+      "learning_rate": 4.675793748323378e-06,
+      "loss": 0.4829,
+      "step": 8399
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5619466381369013,
+      "learning_rate": 4.6747643398961465e-06,
+      "loss": 0.4192,
+      "step": 8400
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.660468492810452,
+      "learning_rate": 4.6737349453131556e-06,
+      "loss": 0.4634,
+      "step": 8401
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7212004286905599,
+      "learning_rate": 4.672705564618228e-06,
+      "loss": 0.5062,
+      "step": 8402
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.571651013072029,
+      "learning_rate": 4.671676197855178e-06,
+      "loss": 0.4592,
+      "step": 8403
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6756887980190536,
+      "learning_rate": 4.670646845067823e-06,
+      "loss": 0.494,
+      "step": 8404
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7638395564295171,
+      "learning_rate": 4.669617506299979e-06,
+      "loss": 0.5114,
+      "step": 8405
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7078517964132525,
+      "learning_rate": 4.668588181595462e-06,
+      "loss": 0.4377,
+      "step": 8406
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8162889833163972,
+      "learning_rate": 4.667558870998088e-06,
+      "loss": 0.5008,
+      "step": 8407
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6307686552880518,
+      "learning_rate": 4.66652957455167e-06,
+      "loss": 0.5279,
+      "step": 8408
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8860161005715478,
+      "learning_rate": 4.665500292300022e-06,
+      "loss": 0.5348,
+      "step": 8409
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.753833515737956,
+      "learning_rate": 4.6644710242869586e-06,
+      "loss": 0.5833,
+      "step": 8410
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7902452914431704,
+      "learning_rate": 4.66344177055629e-06,
+      "loss": 0.5909,
+      "step": 8411
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8910486117109179,
+      "learning_rate": 4.662412531151831e-06,
+      "loss": 0.5373,
+      "step": 8412
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7628166571938368,
+      "learning_rate": 4.661383306117392e-06,
+      "loss": 0.5369,
+      "step": 8413
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8626640001914128,
+      "learning_rate": 4.66035409549678e-06,
+      "loss": 0.5132,
+      "step": 8414
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6707463882512237,
+      "learning_rate": 4.6593248993338125e-06,
+      "loss": 0.5181,
+      "step": 8415
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7840943731201608,
+      "learning_rate": 4.658295717672295e-06,
+      "loss": 0.5049,
+      "step": 8416
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7362141479085723,
+      "learning_rate": 4.657266550556036e-06,
+      "loss": 0.5073,
+      "step": 8417
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7142310809842124,
+      "learning_rate": 4.656237398028846e-06,
+      "loss": 0.5513,
+      "step": 8418
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7563563098539968,
+      "learning_rate": 4.65520826013453e-06,
+      "loss": 0.5116,
+      "step": 8419
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.643873280231301,
+      "learning_rate": 4.654179136916898e-06,
+      "loss": 0.4757,
+      "step": 8420
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6268166627092473,
+      "learning_rate": 4.653150028419754e-06,
+      "loss": 0.4558,
+      "step": 8421
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6315090494745018,
+      "learning_rate": 4.652120934686905e-06,
+      "loss": 0.4817,
+      "step": 8422
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.730675884373918,
+      "learning_rate": 4.651091855762157e-06,
+      "loss": 0.5048,
+      "step": 8423
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9512091654455435,
+      "learning_rate": 4.650062791689313e-06,
+      "loss": 0.4413,
+      "step": 8424
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6870553837144314,
+      "learning_rate": 4.649033742512178e-06,
+      "loss": 0.5166,
+      "step": 8425
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6898670375512296,
+      "learning_rate": 4.648004708274553e-06,
+      "loss": 0.501,
+      "step": 8426
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6166967528623053,
+      "learning_rate": 4.646975689020244e-06,
+      "loss": 0.4447,
+      "step": 8427
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7622041988668076,
+      "learning_rate": 4.645946684793053e-06,
+      "loss": 0.5427,
+      "step": 8428
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9512668191239007,
+      "learning_rate": 4.64491769563678e-06,
+      "loss": 0.5007,
+      "step": 8429
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6550393821112371,
+      "learning_rate": 4.643888721595226e-06,
+      "loss": 0.4389,
+      "step": 8430
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5790761205366833,
+      "learning_rate": 4.64285976271219e-06,
+      "loss": 0.4697,
+      "step": 8431
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7332266616339383,
+      "learning_rate": 4.6418308190314735e-06,
+      "loss": 0.5028,
+      "step": 8432
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7020137781479109,
+      "learning_rate": 4.640801890596875e-06,
+      "loss": 0.4921,
+      "step": 8433
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8762488343565392,
+      "learning_rate": 4.639772977452192e-06,
+      "loss": 0.5679,
+      "step": 8434
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6958711154939297,
+      "learning_rate": 4.638744079641222e-06,
+      "loss": 0.492,
+      "step": 8435
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.689114992115966,
+      "learning_rate": 4.6377151972077616e-06,
+      "loss": 0.4725,
+      "step": 8436
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7472976433703624,
+      "learning_rate": 4.636686330195608e-06,
+      "loss": 0.5361,
+      "step": 8437
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7625279445897821,
+      "learning_rate": 4.635657478648554e-06,
+      "loss": 0.5082,
+      "step": 8438
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7107890464207818,
+      "learning_rate": 4.6346286426104e-06,
+      "loss": 0.5421,
+      "step": 8439
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6283271478914435,
+      "learning_rate": 4.633599822124936e-06,
+      "loss": 0.4799,
+      "step": 8440
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7179127541771869,
+      "learning_rate": 4.632571017235958e-06,
+      "loss": 0.4939,
+      "step": 8441
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5929685599772619,
+      "learning_rate": 4.631542227987256e-06,
+      "loss": 0.4426,
+      "step": 8442
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7619415505757918,
+      "learning_rate": 4.630513454422625e-06,
+      "loss": 0.4697,
+      "step": 8443
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6145496912518909,
+      "learning_rate": 4.629484696585855e-06,
+      "loss": 0.4827,
+      "step": 8444
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7917648199349726,
+      "learning_rate": 4.628455954520737e-06,
+      "loss": 0.5028,
+      "step": 8445
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6429403948703459,
+      "learning_rate": 4.627427228271063e-06,
+      "loss": 0.4956,
+      "step": 8446
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7043783357359542,
+      "learning_rate": 4.626398517880621e-06,
+      "loss": 0.4971,
+      "step": 8447
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5603148106640026,
+      "learning_rate": 4.6253698233932e-06,
+      "loss": 0.4695,
+      "step": 8448
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7000625029904758,
+      "learning_rate": 4.624341144852589e-06,
+      "loss": 0.4331,
+      "step": 8449
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7235475734831217,
+      "learning_rate": 4.623312482302574e-06,
+      "loss": 0.4813,
+      "step": 8450
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7164336906123124,
+      "learning_rate": 4.622283835786942e-06,
+      "loss": 0.4594,
+      "step": 8451
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8073894237025039,
+      "learning_rate": 4.621255205349482e-06,
+      "loss": 0.569,
+      "step": 8452
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8459963819415339,
+      "learning_rate": 4.620226591033977e-06,
+      "loss": 0.5264,
+      "step": 8453
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7778030068358962,
+      "learning_rate": 4.619197992884213e-06,
+      "loss": 0.5397,
+      "step": 8454
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5668433521425842,
+      "learning_rate": 4.618169410943973e-06,
+      "loss": 0.4623,
+      "step": 8455
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6018822436727288,
+      "learning_rate": 4.61714084525704e-06,
+      "loss": 0.4821,
+      "step": 8456
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6914846077201479,
+      "learning_rate": 4.616112295867199e-06,
+      "loss": 0.4795,
+      "step": 8457
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.626801691390319,
+      "learning_rate": 4.615083762818231e-06,
+      "loss": 0.4504,
+      "step": 8458
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5912242521165715,
+      "learning_rate": 4.614055246153916e-06,
+      "loss": 0.4401,
+      "step": 8459
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6375960596102452,
+      "learning_rate": 4.613026745918037e-06,
+      "loss": 0.4583,
+      "step": 8460
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7199161450730038,
+      "learning_rate": 4.611998262154373e-06,
+      "loss": 0.4799,
+      "step": 8461
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7765032892839526,
+      "learning_rate": 4.610969794906703e-06,
+      "loss": 0.5635,
+      "step": 8462
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7003572808161469,
+      "learning_rate": 4.609941344218804e-06,
+      "loss": 0.5241,
+      "step": 8463
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7201438386129125,
+      "learning_rate": 4.608912910134457e-06,
+      "loss": 0.5248,
+      "step": 8464
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.631313299824022,
+      "learning_rate": 4.607884492697437e-06,
+      "loss": 0.4669,
+      "step": 8465
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8139530153427013,
+      "learning_rate": 4.606856091951523e-06,
+      "loss": 0.489,
+      "step": 8466
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5879956968665803,
+      "learning_rate": 4.605827707940488e-06,
+      "loss": 0.4634,
+      "step": 8467
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6685052116960838,
+      "learning_rate": 4.604799340708107e-06,
+      "loss": 0.541,
+      "step": 8468
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7104296776717307,
+      "learning_rate": 4.6037709902981555e-06,
+      "loss": 0.474,
+      "step": 8469
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6014689511117821,
+      "learning_rate": 4.602742656754407e-06,
+      "loss": 0.4802,
+      "step": 8470
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6157666300039946,
+      "learning_rate": 4.601714340120634e-06,
+      "loss": 0.4722,
+      "step": 8471
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.656486953096236,
+      "learning_rate": 4.600686040440609e-06,
+      "loss": 0.4995,
+      "step": 8472
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7721062084650561,
+      "learning_rate": 4.599657757758103e-06,
+      "loss": 0.5929,
+      "step": 8473
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6894633827978588,
+      "learning_rate": 4.598629492116887e-06,
+      "loss": 0.4997,
+      "step": 8474
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8119950004487392,
+      "learning_rate": 4.597601243560731e-06,
+      "loss": 0.5158,
+      "step": 8475
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7332548052297658,
+      "learning_rate": 4.596573012133403e-06,
+      "loss": 0.5482,
+      "step": 8476
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6992240521993527,
+      "learning_rate": 4.595544797878673e-06,
+      "loss": 0.4908,
+      "step": 8477
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7172619869783801,
+      "learning_rate": 4.59451660084031e-06,
+      "loss": 0.4776,
+      "step": 8478
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5937967832691354,
+      "learning_rate": 4.593488421062079e-06,
+      "loss": 0.4863,
+      "step": 8479
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7490510484117048,
+      "learning_rate": 4.592460258587746e-06,
+      "loss": 0.4768,
+      "step": 8480
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6532429268903567,
+      "learning_rate": 4.591432113461079e-06,
+      "loss": 0.447,
+      "step": 8481
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6112587121051494,
+      "learning_rate": 4.59040398572584e-06,
+      "loss": 0.4357,
+      "step": 8482
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7077063572588193,
+      "learning_rate": 4.589375875425795e-06,
+      "loss": 0.4799,
+      "step": 8483
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6327200742886288,
+      "learning_rate": 4.5883477826047075e-06,
+      "loss": 0.4513,
+      "step": 8484
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6118384887184034,
+      "learning_rate": 4.5873197073063385e-06,
+      "loss": 0.428,
+      "step": 8485
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6382769090767949,
+      "learning_rate": 4.586291649574451e-06,
+      "loss": 0.4515,
+      "step": 8486
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.63640148937763,
+      "learning_rate": 4.585263609452807e-06,
+      "loss": 0.5084,
+      "step": 8487
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6005979336395098,
+      "learning_rate": 4.584235586985162e-06,
+      "loss": 0.5446,
+      "step": 8488
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7239166760827016,
+      "learning_rate": 4.583207582215283e-06,
+      "loss": 0.5194,
+      "step": 8489
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6976519505597664,
+      "learning_rate": 4.582179595186925e-06,
+      "loss": 0.5144,
+      "step": 8490
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8564321157553273,
+      "learning_rate": 4.581151625943846e-06,
+      "loss": 0.5568,
+      "step": 8491
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.019726722347934,
+      "learning_rate": 4.5801236745298035e-06,
+      "loss": 0.5156,
+      "step": 8492
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7631858368641522,
+      "learning_rate": 4.579095740988555e-06,
+      "loss": 0.5364,
+      "step": 8493
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7451487220324169,
+      "learning_rate": 4.5780678253638565e-06,
+      "loss": 0.5113,
+      "step": 8494
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5332420172698064,
+      "learning_rate": 4.577039927699461e-06,
+      "loss": 0.3696,
+      "step": 8495
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7023011347967351,
+      "learning_rate": 4.576012048039126e-06,
+      "loss": 0.445,
+      "step": 8496
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5533543310108555,
+      "learning_rate": 4.574984186426602e-06,
+      "loss": 0.4346,
+      "step": 8497
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5938299622887316,
+      "learning_rate": 4.573956342905643e-06,
+      "loss": 0.4728,
+      "step": 8498
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7108504577666083,
+      "learning_rate": 4.572928517520001e-06,
+      "loss": 0.5086,
+      "step": 8499
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7453909158545868,
+      "learning_rate": 4.5719007103134255e-06,
+      "loss": 0.5116,
+      "step": 8500
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.1696602653173944,
+      "learning_rate": 4.570872921329671e-06,
+      "loss": 0.5146,
+      "step": 8501
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.651065938209936,
+      "learning_rate": 4.569845150612485e-06,
+      "loss": 0.5385,
+      "step": 8502
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.783421636126263,
+      "learning_rate": 4.5688173982056155e-06,
+      "loss": 0.4935,
+      "step": 8503
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6527657778411697,
+      "learning_rate": 4.567789664152812e-06,
+      "loss": 0.5161,
+      "step": 8504
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7547590276723879,
+      "learning_rate": 4.566761948497821e-06,
+      "loss": 0.4957,
+      "step": 8505
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6484318284195962,
+      "learning_rate": 4.5657342512843905e-06,
+      "loss": 0.5079,
+      "step": 8506
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6499076906790505,
+      "learning_rate": 4.5647065725562646e-06,
+      "loss": 0.4986,
+      "step": 8507
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6062918986988345,
+      "learning_rate": 4.563678912357189e-06,
+      "loss": 0.4167,
+      "step": 8508
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6235554844187612,
+      "learning_rate": 4.562651270730907e-06,
+      "loss": 0.4319,
+      "step": 8509
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6540367638124763,
+      "learning_rate": 4.561623647721164e-06,
+      "loss": 0.4961,
+      "step": 8510
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.354005505526635,
+      "learning_rate": 4.560596043371701e-06,
+      "loss": 0.57,
+      "step": 8511
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6988677283713708,
+      "learning_rate": 4.55956845772626e-06,
+      "loss": 0.504,
+      "step": 8512
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7861868834231205,
+      "learning_rate": 4.5585408908285816e-06,
+      "loss": 0.5371,
+      "step": 8513
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8069722489047302,
+      "learning_rate": 4.557513342722408e-06,
+      "loss": 0.5275,
+      "step": 8514
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6365879697457588,
+      "learning_rate": 4.5564858134514775e-06,
+      "loss": 0.4497,
+      "step": 8515
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6273633559367239,
+      "learning_rate": 4.555458303059528e-06,
+      "loss": 0.4146,
+      "step": 8516
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8136841886984113,
+      "learning_rate": 4.554430811590298e-06,
+      "loss": 0.5138,
+      "step": 8517
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8264935032692735,
+      "learning_rate": 4.553403339087525e-06,
+      "loss": 0.487,
+      "step": 8518
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.121747488940747,
+      "learning_rate": 4.5523758855949436e-06,
+      "loss": 0.4778,
+      "step": 8519
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.700967586858247,
+      "learning_rate": 4.551348451156291e-06,
+      "loss": 0.492,
+      "step": 8520
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.63724114384012,
+      "learning_rate": 4.550321035815302e-06,
+      "loss": 0.4543,
+      "step": 8521
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6270102790803425,
+      "learning_rate": 4.549293639615709e-06,
+      "loss": 0.4213,
+      "step": 8522
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.582054691314822,
+      "learning_rate": 4.548266262601245e-06,
+      "loss": 0.4826,
+      "step": 8523
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6218453651697858,
+      "learning_rate": 4.5472389048156435e-06,
+      "loss": 0.4312,
+      "step": 8524
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6673240556469603,
+      "learning_rate": 4.5462115663026334e-06,
+      "loss": 0.466,
+      "step": 8525
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7173811025798682,
+      "learning_rate": 4.545184247105948e-06,
+      "loss": 0.5349,
+      "step": 8526
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7503162537633045,
+      "learning_rate": 4.544156947269316e-06,
+      "loss": 0.5353,
+      "step": 8527
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6175185389318006,
+      "learning_rate": 4.543129666836467e-06,
+      "loss": 0.4338,
+      "step": 8528
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5703059141701459,
+      "learning_rate": 4.542102405851127e-06,
+      "loss": 0.4122,
+      "step": 8529
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6690488920865705,
+      "learning_rate": 4.541075164357024e-06,
+      "loss": 0.4688,
+      "step": 8530
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.0316807498732905,
+      "learning_rate": 4.5400479423978855e-06,
+      "loss": 0.4864,
+      "step": 8531
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.04091603041933,
+      "learning_rate": 4.539020740017436e-06,
+      "loss": 0.5029,
+      "step": 8532
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6447024650700228,
+      "learning_rate": 4.5379935572594e-06,
+      "loss": 0.4826,
+      "step": 8533
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7603019123434447,
+      "learning_rate": 4.536966394167503e-06,
+      "loss": 0.5101,
+      "step": 8534
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5531883537578846,
+      "learning_rate": 4.535939250785466e-06,
+      "loss": 0.3976,
+      "step": 8535
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5895226896128738,
+      "learning_rate": 4.534912127157013e-06,
+      "loss": 0.4469,
+      "step": 8536
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8337964471556024,
+      "learning_rate": 4.533885023325864e-06,
+      "loss": 0.5188,
+      "step": 8537
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.796924810287293,
+      "learning_rate": 4.532857939335738e-06,
+      "loss": 0.5579,
+      "step": 8538
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7125114184666226,
+      "learning_rate": 4.531830875230359e-06,
+      "loss": 0.5255,
+      "step": 8539
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7751007418744865,
+      "learning_rate": 4.5308038310534435e-06,
+      "loss": 0.512,
+      "step": 8540
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5971475669635828,
+      "learning_rate": 4.529776806848709e-06,
+      "loss": 0.4673,
+      "step": 8541
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7450481095365744,
+      "learning_rate": 4.528749802659875e-06,
+      "loss": 0.5264,
+      "step": 8542
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6659255352488508,
+      "learning_rate": 4.5277228185306545e-06,
+      "loss": 0.4933,
+      "step": 8543
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.629523510727235,
+      "learning_rate": 4.5266958545047645e-06,
+      "loss": 0.4729,
+      "step": 8544
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8119753801888152,
+      "learning_rate": 4.52566891062592e-06,
+      "loss": 0.4691,
+      "step": 8545
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6606991694064694,
+      "learning_rate": 4.5246419869378335e-06,
+      "loss": 0.4831,
+      "step": 8546
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7158881713966089,
+      "learning_rate": 4.52361508348422e-06,
+      "loss": 0.5312,
+      "step": 8547
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.2598959055127832,
+      "learning_rate": 4.522588200308789e-06,
+      "loss": 0.5205,
+      "step": 8548
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.4816377323870071,
+      "learning_rate": 4.521561337455252e-06,
+      "loss": 0.5223,
+      "step": 8549
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6636007298329161,
+      "learning_rate": 4.520534494967319e-06,
+      "loss": 0.4837,
+      "step": 8550
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.974156257685153,
+      "learning_rate": 4.519507672888703e-06,
+      "loss": 0.5202,
+      "step": 8551
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6543097079647269,
+      "learning_rate": 4.518480871263109e-06,
+      "loss": 0.4973,
+      "step": 8552
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7239561621348768,
+      "learning_rate": 4.517454090134246e-06,
+      "loss": 0.4602,
+      "step": 8553
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6288596350119882,
+      "learning_rate": 4.5164273295458196e-06,
+      "loss": 0.4684,
+      "step": 8554
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.596747871792202,
+      "learning_rate": 4.515400589541537e-06,
+      "loss": 0.4458,
+      "step": 8555
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6388115211221093,
+      "learning_rate": 4.5143738701651035e-06,
+      "loss": 0.453,
+      "step": 8556
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8543492795767097,
+      "learning_rate": 4.5133471714602215e-06,
+      "loss": 0.5549,
+      "step": 8557
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.910159438413612,
+      "learning_rate": 4.512320493470596e-06,
+      "loss": 0.4789,
+      "step": 8558
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5737940554333593,
+      "learning_rate": 4.5112938362399284e-06,
+      "loss": 0.4676,
+      "step": 8559
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7892243399036161,
+      "learning_rate": 4.51026719981192e-06,
+      "loss": 0.5219,
+      "step": 8560
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6176669981869405,
+      "learning_rate": 4.5092405842302725e-06,
+      "loss": 0.4941,
+      "step": 8561
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6739875231464935,
+      "learning_rate": 4.508213989538683e-06,
+      "loss": 0.4539,
+      "step": 8562
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7454513425658681,
+      "learning_rate": 4.507187415780855e-06,
+      "loss": 0.5714,
+      "step": 8563
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.5481377961594522,
+      "learning_rate": 4.506160863000483e-06,
+      "loss": 0.5671,
+      "step": 8564
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.687251456902005,
+      "learning_rate": 4.505134331241265e-06,
+      "loss": 0.4877,
+      "step": 8565
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7327951515727712,
+      "learning_rate": 4.504107820546898e-06,
+      "loss": 0.4892,
+      "step": 8566
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7368247642955208,
+      "learning_rate": 4.503081330961076e-06,
+      "loss": 0.5365,
+      "step": 8567
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8655420527301664,
+      "learning_rate": 4.502054862527494e-06,
+      "loss": 0.5401,
+      "step": 8568
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.668336103426427,
+      "learning_rate": 4.5010284152898445e-06,
+      "loss": 0.487,
+      "step": 8569
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6543585134052369,
+      "learning_rate": 4.500001989291821e-06,
+      "loss": 0.4561,
+      "step": 8570
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.616512254387168,
+      "learning_rate": 4.498975584577116e-06,
+      "loss": 0.4662,
+      "step": 8571
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5417851188518977,
+      "learning_rate": 4.497949201189418e-06,
+      "loss": 0.4544,
+      "step": 8572
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6382641415253936,
+      "learning_rate": 4.49692283917242e-06,
+      "loss": 0.4463,
+      "step": 8573
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6068944434643876,
+      "learning_rate": 4.495896498569807e-06,
+      "loss": 0.4646,
+      "step": 8574
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7079080744809629,
+      "learning_rate": 4.49487017942527e-06,
+      "loss": 0.4764,
+      "step": 8575
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5734771413038712,
+      "learning_rate": 4.493843881782495e-06,
+      "loss": 0.4756,
+      "step": 8576
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6236499671704739,
+      "learning_rate": 4.492817605685169e-06,
+      "loss": 0.4982,
+      "step": 8577
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.9465062605165874,
+      "learning_rate": 4.491791351176977e-06,
+      "loss": 0.5133,
+      "step": 8578
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6049141820933159,
+      "learning_rate": 4.490765118301603e-06,
+      "loss": 0.4948,
+      "step": 8579
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7607790457656651,
+      "learning_rate": 4.48973890710273e-06,
+      "loss": 0.4705,
+      "step": 8580
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.629780419140997,
+      "learning_rate": 4.4887127176240415e-06,
+      "loss": 0.493,
+      "step": 8581
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6243714315478958,
+      "learning_rate": 4.487686549909219e-06,
+      "loss": 0.4483,
+      "step": 8582
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6057055072199797,
+      "learning_rate": 4.486660404001944e-06,
+      "loss": 0.4305,
+      "step": 8583
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6615294020976404,
+      "learning_rate": 4.485634279945894e-06,
+      "loss": 0.4765,
+      "step": 8584
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6795424287282442,
+      "learning_rate": 4.484608177784749e-06,
+      "loss": 0.4713,
+      "step": 8585
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5962094773294149,
+      "learning_rate": 4.483582097562187e-06,
+      "loss": 0.4787,
+      "step": 8586
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7227475521486151,
+      "learning_rate": 4.482556039321884e-06,
+      "loss": 0.4962,
+      "step": 8587
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6434688661805869,
+      "learning_rate": 4.481530003107518e-06,
+      "loss": 0.4598,
+      "step": 8588
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5934801999275032,
+      "learning_rate": 4.480503988962762e-06,
+      "loss": 0.4845,
+      "step": 8589
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.753699827132346,
+      "learning_rate": 4.479477996931293e-06,
+      "loss": 0.57,
+      "step": 8590
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7304870421957352,
+      "learning_rate": 4.478452027056781e-06,
+      "loss": 0.5929,
+      "step": 8591
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8228141212036288,
+      "learning_rate": 4.4774260793829e-06,
+      "loss": 0.5648,
+      "step": 8592
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5714437996989518,
+      "learning_rate": 4.476400153953321e-06,
+      "loss": 0.4417,
+      "step": 8593
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6184048249245075,
+      "learning_rate": 4.475374250811714e-06,
+      "loss": 0.4568,
+      "step": 8594
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.772098798074029,
+      "learning_rate": 4.474348370001749e-06,
+      "loss": 0.5371,
+      "step": 8595
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.9233128394480612,
+      "learning_rate": 4.473322511567094e-06,
+      "loss": 0.518,
+      "step": 8596
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7172785458518779,
+      "learning_rate": 4.472296675551416e-06,
+      "loss": 0.4532,
+      "step": 8597
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7916605861801453,
+      "learning_rate": 4.471270861998383e-06,
+      "loss": 0.542,
+      "step": 8598
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5829875274529767,
+      "learning_rate": 4.4702450709516595e-06,
+      "loss": 0.4628,
+      "step": 8599
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7499182173210165,
+      "learning_rate": 4.4692193024549095e-06,
+      "loss": 0.4486,
+      "step": 8600
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7169296975442507,
+      "learning_rate": 4.468193556551798e-06,
+      "loss": 0.4398,
+      "step": 8601
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6425119022213079,
+      "learning_rate": 4.467167833285988e-06,
+      "loss": 0.4515,
+      "step": 8602
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5763835112707627,
+      "learning_rate": 4.466142132701141e-06,
+      "loss": 0.4667,
+      "step": 8603
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6296111599683203,
+      "learning_rate": 4.465116454840918e-06,
+      "loss": 0.5095,
+      "step": 8604
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.747651229603383,
+      "learning_rate": 4.4640907997489765e-06,
+      "loss": 0.5434,
+      "step": 8605
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7213999981081614,
+      "learning_rate": 4.4630651674689785e-06,
+      "loss": 0.4558,
+      "step": 8606
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7039531633040491,
+      "learning_rate": 4.4620395580445806e-06,
+      "loss": 0.48,
+      "step": 8607
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5914858093262314,
+      "learning_rate": 4.461013971519439e-06,
+      "loss": 0.4814,
+      "step": 8608
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7893659125193687,
+      "learning_rate": 4.45998840793721e-06,
+      "loss": 0.584,
+      "step": 8609
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8114567157705281,
+      "learning_rate": 4.45896286734155e-06,
+      "loss": 0.4689,
+      "step": 8610
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6929330862075997,
+      "learning_rate": 4.457937349776111e-06,
+      "loss": 0.4272,
+      "step": 8611
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7801562241755182,
+      "learning_rate": 4.456911855284545e-06,
+      "loss": 0.4953,
+      "step": 8612
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6751757870398382,
+      "learning_rate": 4.455886383910507e-06,
+      "loss": 0.5268,
+      "step": 8613
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7973034526726619,
+      "learning_rate": 4.454860935697648e-06,
+      "loss": 0.5348,
+      "step": 8614
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6372102847295429,
+      "learning_rate": 4.453835510689617e-06,
+      "loss": 0.4681,
+      "step": 8615
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6916366072561504,
+      "learning_rate": 4.452810108930063e-06,
+      "loss": 0.5278,
+      "step": 8616
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8292440576209625,
+      "learning_rate": 4.451784730462634e-06,
+      "loss": 0.5617,
+      "step": 8617
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6891838474682641,
+      "learning_rate": 4.450759375330977e-06,
+      "loss": 0.5236,
+      "step": 8618
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6294819532266165,
+      "learning_rate": 4.4497340435787385e-06,
+      "loss": 0.4739,
+      "step": 8619
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6667224319215789,
+      "learning_rate": 4.4487087352495625e-06,
+      "loss": 0.475,
+      "step": 8620
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6618942768878272,
+      "learning_rate": 4.447683450387095e-06,
+      "loss": 0.4744,
+      "step": 8621
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7378393210761782,
+      "learning_rate": 4.446658189034977e-06,
+      "loss": 0.5222,
+      "step": 8622
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6726329303074522,
+      "learning_rate": 4.4456329512368525e-06,
+      "loss": 0.4672,
+      "step": 8623
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7723465303164209,
+      "learning_rate": 4.44460773703636e-06,
+      "loss": 0.5219,
+      "step": 8624
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6935111887510977,
+      "learning_rate": 4.443582546477143e-06,
+      "loss": 0.4762,
+      "step": 8625
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.565089202783496,
+      "learning_rate": 4.4425573796028385e-06,
+      "loss": 0.4188,
+      "step": 8626
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5791298537921312,
+      "learning_rate": 4.4415322364570854e-06,
+      "loss": 0.4689,
+      "step": 8627
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6218219346859978,
+      "learning_rate": 4.44050711708352e-06,
+      "loss": 0.4506,
+      "step": 8628
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6059194126133071,
+      "learning_rate": 4.43948202152578e-06,
+      "loss": 0.4636,
+      "step": 8629
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5934545801783437,
+      "learning_rate": 4.438456949827498e-06,
+      "loss": 0.4531,
+      "step": 8630
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6532553877571453,
+      "learning_rate": 4.43743190203231e-06,
+      "loss": 0.4845,
+      "step": 8631
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7466083408297207,
+      "learning_rate": 4.436406878183848e-06,
+      "loss": 0.5414,
+      "step": 8632
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7449060202951123,
+      "learning_rate": 4.435381878325744e-06,
+      "loss": 0.526,
+      "step": 8633
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6527145089285332,
+      "learning_rate": 4.434356902501629e-06,
+      "loss": 0.5192,
+      "step": 8634
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5614537143279134,
+      "learning_rate": 4.433331950755133e-06,
+      "loss": 0.4947,
+      "step": 8635
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6354571424910672,
+      "learning_rate": 4.432307023129885e-06,
+      "loss": 0.4495,
+      "step": 8636
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6750311823057951,
+      "learning_rate": 4.431282119669513e-06,
+      "loss": 0.4375,
+      "step": 8637
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6404922351551362,
+      "learning_rate": 4.430257240417644e-06,
+      "loss": 0.4317,
+      "step": 8638
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5444893108505425,
+      "learning_rate": 4.429232385417903e-06,
+      "loss": 0.3925,
+      "step": 8639
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.61761876384027,
+      "learning_rate": 4.428207554713916e-06,
+      "loss": 0.4815,
+      "step": 8640
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7125928483156411,
+      "learning_rate": 4.4271827483493065e-06,
+      "loss": 0.5078,
+      "step": 8641
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7224533202582047,
+      "learning_rate": 4.4261579663676964e-06,
+      "loss": 0.5342,
+      "step": 8642
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8157519984574083,
+      "learning_rate": 4.425133208812708e-06,
+      "loss": 0.5587,
+      "step": 8643
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.9054439381780619,
+      "learning_rate": 4.424108475727962e-06,
+      "loss": 0.5551,
+      "step": 8644
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6051133046357489,
+      "learning_rate": 4.423083767157078e-06,
+      "loss": 0.4632,
+      "step": 8645
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6810378920543833,
+      "learning_rate": 4.422059083143675e-06,
+      "loss": 0.5001,
+      "step": 8646
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8214147314192999,
+      "learning_rate": 4.4210344237313695e-06,
+      "loss": 0.5469,
+      "step": 8647
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7436719018597728,
+      "learning_rate": 4.420009788963779e-06,
+      "loss": 0.5448,
+      "step": 8648
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6598472872452492,
+      "learning_rate": 4.4189851788845165e-06,
+      "loss": 0.5042,
+      "step": 8649
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.9718468638325388,
+      "learning_rate": 4.4179605935372e-06,
+      "loss": 0.47,
+      "step": 8650
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5969621057466895,
+      "learning_rate": 4.41693603296544e-06,
+      "loss": 0.453,
+      "step": 8651
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5265999909546717,
+      "learning_rate": 4.415911497212852e-06,
+      "loss": 0.394,
+      "step": 8652
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6303210903993633,
+      "learning_rate": 4.414886986323043e-06,
+      "loss": 0.4368,
+      "step": 8653
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6939423926267622,
+      "learning_rate": 4.4138625003396265e-06,
+      "loss": 0.4152,
+      "step": 8654
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.626467544574878,
+      "learning_rate": 4.41283803930621e-06,
+      "loss": 0.5155,
+      "step": 8655
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7689550372434832,
+      "learning_rate": 4.411813603266401e-06,
+      "loss": 0.4649,
+      "step": 8656
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6866838720873619,
+      "learning_rate": 4.410789192263808e-06,
+      "loss": 0.526,
+      "step": 8657
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8654830526462901,
+      "learning_rate": 4.409764806342037e-06,
+      "loss": 0.5443,
+      "step": 8658
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5976793751090987,
+      "learning_rate": 4.408740445544691e-06,
+      "loss": 0.4873,
+      "step": 8659
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5784852569203252,
+      "learning_rate": 4.407716109915375e-06,
+      "loss": 0.4481,
+      "step": 8660
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6564506025264064,
+      "learning_rate": 4.406691799497692e-06,
+      "loss": 0.4941,
+      "step": 8661
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7154700971699651,
+      "learning_rate": 4.405667514335241e-06,
+      "loss": 0.5411,
+      "step": 8662
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6501901165115125,
+      "learning_rate": 4.404643254471626e-06,
+      "loss": 0.4738,
+      "step": 8663
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6076375323665985,
+      "learning_rate": 4.403619019950446e-06,
+      "loss": 0.4509,
+      "step": 8664
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6591207461291212,
+      "learning_rate": 4.4025948108152985e-06,
+      "loss": 0.4859,
+      "step": 8665
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6272720824264322,
+      "learning_rate": 4.401570627109781e-06,
+      "loss": 0.5301,
+      "step": 8666
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6913948497572918,
+      "learning_rate": 4.400546468877489e-06,
+      "loss": 0.4798,
+      "step": 8667
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6434516985941561,
+      "learning_rate": 4.3995223361620185e-06,
+      "loss": 0.4689,
+      "step": 8668
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5876805938368626,
+      "learning_rate": 4.398498229006963e-06,
+      "loss": 0.4696,
+      "step": 8669
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6057948307165987,
+      "learning_rate": 4.3974741474559165e-06,
+      "loss": 0.4372,
+      "step": 8670
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6715480883261464,
+      "learning_rate": 4.39645009155247e-06,
+      "loss": 0.456,
+      "step": 8671
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7285974130030384,
+      "learning_rate": 4.395426061340214e-06,
+      "loss": 0.5578,
+      "step": 8672
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6896630406295814,
+      "learning_rate": 4.394402056862739e-06,
+      "loss": 0.4578,
+      "step": 8673
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6618818964229133,
+      "learning_rate": 4.39337807816363e-06,
+      "loss": 0.4832,
+      "step": 8674
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7675535119380845,
+      "learning_rate": 4.392354125286482e-06,
+      "loss": 0.5144,
+      "step": 8675
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8639253036233522,
+      "learning_rate": 4.391330198274875e-06,
+      "loss": 0.4811,
+      "step": 8676
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6290323980018491,
+      "learning_rate": 4.390306297172398e-06,
+      "loss": 0.4495,
+      "step": 8677
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5589253284305435,
+      "learning_rate": 4.389282422022632e-06,
+      "loss": 0.4362,
+      "step": 8678
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6243316474311928,
+      "learning_rate": 4.388258572869163e-06,
+      "loss": 0.4681,
+      "step": 8679
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7325730737097648,
+      "learning_rate": 4.387234749755572e-06,
+      "loss": 0.486,
+      "step": 8680
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6590763989928909,
+      "learning_rate": 4.386210952725438e-06,
+      "loss": 0.4609,
+      "step": 8681
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6050061783622741,
+      "learning_rate": 4.3851871818223436e-06,
+      "loss": 0.4959,
+      "step": 8682
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8861113074397442,
+      "learning_rate": 4.3841634370898665e-06,
+      "loss": 0.5099,
+      "step": 8683
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6315380546199132,
+      "learning_rate": 4.383139718571583e-06,
+      "loss": 0.4538,
+      "step": 8684
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7651877298522781,
+      "learning_rate": 4.3821160263110706e-06,
+      "loss": 0.5382,
+      "step": 8685
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8077359680707338,
+      "learning_rate": 4.381092360351903e-06,
+      "loss": 0.5179,
+      "step": 8686
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6692613847095965,
+      "learning_rate": 4.380068720737657e-06,
+      "loss": 0.4142,
+      "step": 8687
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6390244607802591,
+      "learning_rate": 4.379045107511905e-06,
+      "loss": 0.4423,
+      "step": 8688
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5724714658015041,
+      "learning_rate": 4.378021520718219e-06,
+      "loss": 0.4826,
+      "step": 8689
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6262749841077924,
+      "learning_rate": 4.376997960400169e-06,
+      "loss": 0.472,
+      "step": 8690
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6323066172468694,
+      "learning_rate": 4.375974426601325e-06,
+      "loss": 0.4979,
+      "step": 8691
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7704174814894261,
+      "learning_rate": 4.374950919365256e-06,
+      "loss": 0.5666,
+      "step": 8692
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6165845872604485,
+      "learning_rate": 4.373927438735528e-06,
+      "loss": 0.4665,
+      "step": 8693
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7284894634413241,
+      "learning_rate": 4.3729039847557085e-06,
+      "loss": 0.4949,
+      "step": 8694
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7551244400144178,
+      "learning_rate": 4.371880557469363e-06,
+      "loss": 0.5939,
+      "step": 8695
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7800292378269323,
+      "learning_rate": 4.370857156920055e-06,
+      "loss": 0.5347,
+      "step": 8696
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7780927129731123,
+      "learning_rate": 4.369833783151347e-06,
+      "loss": 0.5968,
+      "step": 8697
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.766192995022262,
+      "learning_rate": 4.368810436206801e-06,
+      "loss": 0.5426,
+      "step": 8698
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.818234690258935,
+      "learning_rate": 4.3677871161299766e-06,
+      "loss": 0.5855,
+      "step": 8699
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7936641865945461,
+      "learning_rate": 4.366763822964436e-06,
+      "loss": 0.5746,
+      "step": 8700
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7858493847198313,
+      "learning_rate": 4.365740556753736e-06,
+      "loss": 0.531,
+      "step": 8701
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6234169750185313,
+      "learning_rate": 4.364717317541433e-06,
+      "loss": 0.4522,
+      "step": 8702
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5651259641054174,
+      "learning_rate": 4.363694105371085e-06,
+      "loss": 0.4544,
+      "step": 8703
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.714957264872062,
+      "learning_rate": 4.362670920286243e-06,
+      "loss": 0.4892,
+      "step": 8704
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6221203498088445,
+      "learning_rate": 4.361647762330465e-06,
+      "loss": 0.4668,
+      "step": 8705
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8096269206852756,
+      "learning_rate": 4.360624631547301e-06,
+      "loss": 0.4826,
+      "step": 8706
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7454281668502041,
+      "learning_rate": 4.359601527980305e-06,
+      "loss": 0.4689,
+      "step": 8707
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.623672530163333,
+      "learning_rate": 4.358578451673024e-06,
+      "loss": 0.4782,
+      "step": 8708
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6512129741626779,
+      "learning_rate": 4.35755540266901e-06,
+      "loss": 0.4442,
+      "step": 8709
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6338054233723615,
+      "learning_rate": 4.356532381011809e-06,
+      "loss": 0.444,
+      "step": 8710
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6410156350100498,
+      "learning_rate": 4.355509386744967e-06,
+      "loss": 0.4419,
+      "step": 8711
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8272582864463619,
+      "learning_rate": 4.354486419912032e-06,
+      "loss": 0.555,
+      "step": 8712
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.0136044771380013,
+      "learning_rate": 4.353463480556548e-06,
+      "loss": 0.5477,
+      "step": 8713
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6316251371016259,
+      "learning_rate": 4.3524405687220564e-06,
+      "loss": 0.457,
+      "step": 8714
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5959454485535449,
+      "learning_rate": 4.351417684452101e-06,
+      "loss": 0.4442,
+      "step": 8715
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7357473253675132,
+      "learning_rate": 4.350394827790222e-06,
+      "loss": 0.4624,
+      "step": 8716
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.708131638468915,
+      "learning_rate": 4.3493719987799595e-06,
+      "loss": 0.494,
+      "step": 8717
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7470162034134946,
+      "learning_rate": 4.3483491974648525e-06,
+      "loss": 0.5011,
+      "step": 8718
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7863145203388021,
+      "learning_rate": 4.347326423888438e-06,
+      "loss": 0.5582,
+      "step": 8719
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7020088784737551,
+      "learning_rate": 4.34630367809425e-06,
+      "loss": 0.4847,
+      "step": 8720
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6526452735990081,
+      "learning_rate": 4.345280960125828e-06,
+      "loss": 0.5078,
+      "step": 8721
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.601958441186249,
+      "learning_rate": 4.344258270026702e-06,
+      "loss": 0.5041,
+      "step": 8722
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6923037317230399,
+      "learning_rate": 4.343235607840406e-06,
+      "loss": 0.5121,
+      "step": 8723
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5976230137798745,
+      "learning_rate": 4.342212973610469e-06,
+      "loss": 0.4259,
+      "step": 8724
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7412289909548726,
+      "learning_rate": 4.3411903673804266e-06,
+      "loss": 0.4473,
+      "step": 8725
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.684473919796735,
+      "learning_rate": 4.340167789193805e-06,
+      "loss": 0.5068,
+      "step": 8726
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6539044304974383,
+      "learning_rate": 4.339145239094132e-06,
+      "loss": 0.4859,
+      "step": 8727
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8439199576465645,
+      "learning_rate": 4.338122717124934e-06,
+      "loss": 0.5716,
+      "step": 8728
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6855460364648731,
+      "learning_rate": 4.337100223329737e-06,
+      "loss": 0.5111,
+      "step": 8729
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6677128874231376,
+      "learning_rate": 4.336077757752066e-06,
+      "loss": 0.5386,
+      "step": 8730
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8486124846884892,
+      "learning_rate": 4.335055320435442e-06,
+      "loss": 0.5168,
+      "step": 8731
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8376220912232718,
+      "learning_rate": 4.3340329114233886e-06,
+      "loss": 0.4783,
+      "step": 8732
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6469251931162432,
+      "learning_rate": 4.333010530759426e-06,
+      "loss": 0.4524,
+      "step": 8733
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5702683842430247,
+      "learning_rate": 4.331988178487074e-06,
+      "loss": 0.4402,
+      "step": 8734
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6489907106342068,
+      "learning_rate": 4.330965854649851e-06,
+      "loss": 0.4388,
+      "step": 8735
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6926037390772614,
+      "learning_rate": 4.32994355929127e-06,
+      "loss": 0.4754,
+      "step": 8736
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8381220742630461,
+      "learning_rate": 4.328921292454854e-06,
+      "loss": 0.5717,
+      "step": 8737
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6811558963561367,
+      "learning_rate": 4.3278990541841135e-06,
+      "loss": 0.5004,
+      "step": 8738
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8772869906540514,
+      "learning_rate": 4.326876844522563e-06,
+      "loss": 0.5307,
+      "step": 8739
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6755541977106135,
+      "learning_rate": 4.3258546635137135e-06,
+      "loss": 0.5776,
+      "step": 8740
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7146363999171752,
+      "learning_rate": 4.3248325112010775e-06,
+      "loss": 0.5548,
+      "step": 8741
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7573293702064998,
+      "learning_rate": 4.3238103876281635e-06,
+      "loss": 0.5409,
+      "step": 8742
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7290973537859432,
+      "learning_rate": 4.32278829283848e-06,
+      "loss": 0.5383,
+      "step": 8743
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.4555408990939345,
+      "learning_rate": 4.321766226875536e-06,
+      "loss": 0.4785,
+      "step": 8744
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.788367839865495,
+      "learning_rate": 4.320744189782836e-06,
+      "loss": 0.5144,
+      "step": 8745
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.58756658494428,
+      "learning_rate": 4.319722181603885e-06,
+      "loss": 0.4537,
+      "step": 8746
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7559035060987543,
+      "learning_rate": 4.3187002023821875e-06,
+      "loss": 0.4853,
+      "step": 8747
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8002218078278686,
+      "learning_rate": 4.317678252161242e-06,
+      "loss": 0.5377,
+      "step": 8748
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6308985445671924,
+      "learning_rate": 4.316656330984556e-06,
+      "loss": 0.5428,
+      "step": 8749
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7165376891723766,
+      "learning_rate": 4.315634438895626e-06,
+      "loss": 0.4991,
+      "step": 8750
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.762359985958814,
+      "learning_rate": 4.314612575937951e-06,
+      "loss": 0.5207,
+      "step": 8751
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.727014580873897,
+      "learning_rate": 4.313590742155028e-06,
+      "loss": 0.4511,
+      "step": 8752
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8891728874339988,
+      "learning_rate": 4.312568937590354e-06,
+      "loss": 0.5493,
+      "step": 8753
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7236852647320612,
+      "learning_rate": 4.311547162287424e-06,
+      "loss": 0.4813,
+      "step": 8754
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7870785375394218,
+      "learning_rate": 4.3105254162897305e-06,
+      "loss": 0.5121,
+      "step": 8755
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6218075671455104,
+      "learning_rate": 4.309503699640768e-06,
+      "loss": 0.473,
+      "step": 8756
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7602237625412849,
+      "learning_rate": 4.308482012384025e-06,
+      "loss": 0.5028,
+      "step": 8757
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.754321941869743,
+      "learning_rate": 4.307460354562995e-06,
+      "loss": 0.4829,
+      "step": 8758
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5843259666928085,
+      "learning_rate": 4.306438726221163e-06,
+      "loss": 0.4733,
+      "step": 8759
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6226233407243121,
+      "learning_rate": 4.305417127402018e-06,
+      "loss": 0.4648,
+      "step": 8760
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5072978089465365,
+      "learning_rate": 4.304395558149046e-06,
+      "loss": 0.3804,
+      "step": 8761
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5761660222577973,
+      "learning_rate": 4.3033740185057335e-06,
+      "loss": 0.451,
+      "step": 8762
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5597541850739801,
+      "learning_rate": 4.3023525085155625e-06,
+      "loss": 0.458,
+      "step": 8763
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.666625679379583,
+      "learning_rate": 4.301331028222016e-06,
+      "loss": 0.4437,
+      "step": 8764
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6442802108456889,
+      "learning_rate": 4.300309577668575e-06,
+      "loss": 0.4981,
+      "step": 8765
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5518759462909568,
+      "learning_rate": 4.2992881568987175e-06,
+      "loss": 0.4141,
+      "step": 8766
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7275051122919908,
+      "learning_rate": 4.298266765955926e-06,
+      "loss": 0.4797,
+      "step": 8767
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.826745839680901,
+      "learning_rate": 4.297245404883676e-06,
+      "loss": 0.6154,
+      "step": 8768
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7263518334293324,
+      "learning_rate": 4.296224073725443e-06,
+      "loss": 0.52,
+      "step": 8769
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5940595349038942,
+      "learning_rate": 4.2952027725247025e-06,
+      "loss": 0.4792,
+      "step": 8770
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7666531327699513,
+      "learning_rate": 4.294181501324928e-06,
+      "loss": 0.5156,
+      "step": 8771
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5711495507530732,
+      "learning_rate": 4.293160260169591e-06,
+      "loss": 0.477,
+      "step": 8772
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6589047347320118,
+      "learning_rate": 4.292139049102162e-06,
+      "loss": 0.4651,
+      "step": 8773
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5847462779913895,
+      "learning_rate": 4.291117868166113e-06,
+      "loss": 0.5123,
+      "step": 8774
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7016365497518163,
+      "learning_rate": 4.29009671740491e-06,
+      "loss": 0.5135,
+      "step": 8775
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.74615706784095,
+      "learning_rate": 4.289075596862022e-06,
+      "loss": 0.4529,
+      "step": 8776
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8234423431622463,
+      "learning_rate": 4.288054506580914e-06,
+      "loss": 0.5127,
+      "step": 8777
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7751321035269879,
+      "learning_rate": 4.287033446605051e-06,
+      "loss": 0.5716,
+      "step": 8778
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8007655043186032,
+      "learning_rate": 4.2860124169778964e-06,
+      "loss": 0.5661,
+      "step": 8779
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5711652050877859,
+      "learning_rate": 4.28499141774291e-06,
+      "loss": 0.4334,
+      "step": 8780
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7716805471039628,
+      "learning_rate": 4.283970448943555e-06,
+      "loss": 0.4357,
+      "step": 8781
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5555181498804138,
+      "learning_rate": 4.2829495106232895e-06,
+      "loss": 0.425,
+      "step": 8782
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5868305371839402,
+      "learning_rate": 4.281928602825573e-06,
+      "loss": 0.4803,
+      "step": 8783
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6788657402725402,
+      "learning_rate": 4.280907725593861e-06,
+      "loss": 0.4927,
+      "step": 8784
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7123124466256725,
+      "learning_rate": 4.279886878971608e-06,
+      "loss": 0.5404,
+      "step": 8785
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7126931129527746,
+      "learning_rate": 4.278866063002268e-06,
+      "loss": 0.5243,
+      "step": 8786
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6864404507368663,
+      "learning_rate": 4.277845277729298e-06,
+      "loss": 0.5005,
+      "step": 8787
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7678498373500076,
+      "learning_rate": 4.276824523196146e-06,
+      "loss": 0.5279,
+      "step": 8788
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6298741301068632,
+      "learning_rate": 4.275803799446263e-06,
+      "loss": 0.4684,
+      "step": 8789
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8026094621604701,
+      "learning_rate": 4.274783106523099e-06,
+      "loss": 0.573,
+      "step": 8790
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.65629232752146,
+      "learning_rate": 4.2737624444701e-06,
+      "loss": 0.5351,
+      "step": 8791
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8464828505114691,
+      "learning_rate": 4.272741813330713e-06,
+      "loss": 0.5553,
+      "step": 8792
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.756340470284226,
+      "learning_rate": 4.271721213148384e-06,
+      "loss": 0.5518,
+      "step": 8793
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7783301469637252,
+      "learning_rate": 4.270700643966555e-06,
+      "loss": 0.5199,
+      "step": 8794
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6560574216946661,
+      "learning_rate": 4.2696801058286685e-06,
+      "loss": 0.4563,
+      "step": 8795
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5899349082505061,
+      "learning_rate": 4.268659598778166e-06,
+      "loss": 0.4473,
+      "step": 8796
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6963688927216121,
+      "learning_rate": 4.267639122858488e-06,
+      "loss": 0.4467,
+      "step": 8797
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6126800030936337,
+      "learning_rate": 4.2666186781130706e-06,
+      "loss": 0.4689,
+      "step": 8798
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7855198376686179,
+      "learning_rate": 4.265598264585355e-06,
+      "loss": 0.5468,
+      "step": 8799
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.9390285293092171,
+      "learning_rate": 4.264577882318773e-06,
+      "loss": 0.5517,
+      "step": 8800
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6855521345250943,
+      "learning_rate": 4.263557531356761e-06,
+      "loss": 0.4548,
+      "step": 8801
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5769063696116823,
+      "learning_rate": 4.262537211742752e-06,
+      "loss": 0.4425,
+      "step": 8802
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5732954391046863,
+      "learning_rate": 4.261516923520177e-06,
+      "loss": 0.4125,
+      "step": 8803
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5956499422494753,
+      "learning_rate": 4.260496666732466e-06,
+      "loss": 0.4549,
+      "step": 8804
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6353305042450674,
+      "learning_rate": 4.25947644142305e-06,
+      "loss": 0.4711,
+      "step": 8805
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.9048638763083205,
+      "learning_rate": 4.2584562476353545e-06,
+      "loss": 0.5383,
+      "step": 8806
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7014172119541553,
+      "learning_rate": 4.257436085412807e-06,
+      "loss": 0.5191,
+      "step": 8807
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.9372793001861245,
+      "learning_rate": 4.256415954798833e-06,
+      "loss": 0.4708,
+      "step": 8808
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6538557046016998,
+      "learning_rate": 4.255395855836856e-06,
+      "loss": 0.442,
+      "step": 8809
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6381707019285497,
+      "learning_rate": 4.2543757885702956e-06,
+      "loss": 0.4181,
+      "step": 8810
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.580561300876263,
+      "learning_rate": 4.253355753042576e-06,
+      "loss": 0.4428,
+      "step": 8811
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5893239744886439,
+      "learning_rate": 4.252335749297117e-06,
+      "loss": 0.4486,
+      "step": 8812
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6898567139437685,
+      "learning_rate": 4.2513157773773365e-06,
+      "loss": 0.4652,
+      "step": 8813
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7056015871467056,
+      "learning_rate": 4.2502958373266504e-06,
+      "loss": 0.5056,
+      "step": 8814
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5974202773834117,
+      "learning_rate": 4.249275929188475e-06,
+      "loss": 0.4249,
+      "step": 8815
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6334414090393656,
+      "learning_rate": 4.248256053006224e-06,
+      "loss": 0.4754,
+      "step": 8816
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7799495036823637,
+      "learning_rate": 4.247236208823311e-06,
+      "loss": 0.4843,
+      "step": 8817
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5887747793367608,
+      "learning_rate": 4.246216396683147e-06,
+      "loss": 0.527,
+      "step": 8818
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.3043297308638795,
+      "learning_rate": 4.245196616629142e-06,
+      "loss": 0.546,
+      "step": 8819
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8643685450581178,
+      "learning_rate": 4.2441768687047055e-06,
+      "loss": 0.5224,
+      "step": 8820
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6272675124275903,
+      "learning_rate": 4.243157152953244e-06,
+      "loss": 0.481,
+      "step": 8821
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.1797957295211192,
+      "learning_rate": 4.242137469418164e-06,
+      "loss": 0.4391,
+      "step": 8822
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.0286094590805708,
+      "learning_rate": 4.241117818142869e-06,
+      "loss": 0.5224,
+      "step": 8823
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7572385102685983,
+      "learning_rate": 4.2400981991707654e-06,
+      "loss": 0.4413,
+      "step": 8824
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6005078606790096,
+      "learning_rate": 4.239078612545252e-06,
+      "loss": 0.4571,
+      "step": 8825
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6661899109138069,
+      "learning_rate": 4.238059058309731e-06,
+      "loss": 0.444,
+      "step": 8826
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8352314843219187,
+      "learning_rate": 4.2370395365076e-06,
+      "loss": 0.5569,
+      "step": 8827
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.709840176708013,
+      "learning_rate": 4.236020047182259e-06,
+      "loss": 0.5002,
+      "step": 8828
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5991316827714752,
+      "learning_rate": 4.235000590377103e-06,
+      "loss": 0.4729,
+      "step": 8829
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5708281681060035,
+      "learning_rate": 4.233981166135527e-06,
+      "loss": 0.456,
+      "step": 8830
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8995247446030765,
+      "learning_rate": 4.2329617745009255e-06,
+      "loss": 0.523,
+      "step": 8831
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6005560682566539,
+      "learning_rate": 4.2319424155166896e-06,
+      "loss": 0.5023,
+      "step": 8832
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6794776933384548,
+      "learning_rate": 4.230923089226212e-06,
+      "loss": 0.4945,
+      "step": 8833
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8680528274323568,
+      "learning_rate": 4.22990379567288e-06,
+      "loss": 0.5473,
+      "step": 8834
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7804034149022726,
+      "learning_rate": 4.22888453490008e-06,
+      "loss": 0.5322,
+      "step": 8835
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.2199721106463128,
+      "learning_rate": 4.227865306951205e-06,
+      "loss": 0.4914,
+      "step": 8836
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7822910630972447,
+      "learning_rate": 4.226846111869634e-06,
+      "loss": 0.515,
+      "step": 8837
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5832904899114207,
+      "learning_rate": 4.225826949698756e-06,
+      "loss": 0.4443,
+      "step": 8838
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7241557654056496,
+      "learning_rate": 4.224807820481949e-06,
+      "loss": 0.4813,
+      "step": 8839
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7094004004828398,
+      "learning_rate": 4.223788724262597e-06,
+      "loss": 0.5581,
+      "step": 8840
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7708877712455543,
+      "learning_rate": 4.222769661084078e-06,
+      "loss": 0.5518,
+      "step": 8841
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6136329442266498,
+      "learning_rate": 4.22175063098977e-06,
+      "loss": 0.4447,
+      "step": 8842
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6334621032670936,
+      "learning_rate": 4.2207316340230514e-06,
+      "loss": 0.4978,
+      "step": 8843
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6974682168905553,
+      "learning_rate": 4.219712670227297e-06,
+      "loss": 0.4957,
+      "step": 8844
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6308555769922878,
+      "learning_rate": 4.218693739645881e-06,
+      "loss": 0.4816,
+      "step": 8845
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7268561989071667,
+      "learning_rate": 4.217674842322176e-06,
+      "loss": 0.5463,
+      "step": 8846
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5778709276870324,
+      "learning_rate": 4.216655978299552e-06,
+      "loss": 0.4358,
+      "step": 8847
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5998293710339382,
+      "learning_rate": 4.215637147621378e-06,
+      "loss": 0.4907,
+      "step": 8848
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.719870876190144,
+      "learning_rate": 4.214618350331027e-06,
+      "loss": 0.5108,
+      "step": 8849
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7255538086626865,
+      "learning_rate": 4.213599586471863e-06,
+      "loss": 0.5299,
+      "step": 8850
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6374626330450108,
+      "learning_rate": 4.212580856087251e-06,
+      "loss": 0.451,
+      "step": 8851
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7172841179198404,
+      "learning_rate": 4.2115621592205565e-06,
+      "loss": 0.5047,
+      "step": 8852
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8912355896496387,
+      "learning_rate": 4.210543495915141e-06,
+      "loss": 0.5374,
+      "step": 8853
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6367142068190395,
+      "learning_rate": 4.2095248662143666e-06,
+      "loss": 0.4474,
+      "step": 8854
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6315193094392435,
+      "learning_rate": 4.2085062701615916e-06,
+      "loss": 0.4302,
+      "step": 8855
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7011549418738294,
+      "learning_rate": 4.207487707800176e-06,
+      "loss": 0.4922,
+      "step": 8856
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6902072334270198,
+      "learning_rate": 4.206469179173476e-06,
+      "loss": 0.4887,
+      "step": 8857
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5876436355786838,
+      "learning_rate": 4.205450684324847e-06,
+      "loss": 0.4639,
+      "step": 8858
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5818643807679521,
+      "learning_rate": 4.2044322232976444e-06,
+      "loss": 0.4423,
+      "step": 8859
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.568966457609983,
+      "learning_rate": 4.203413796135217e-06,
+      "loss": 0.4479,
+      "step": 8860
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6786900564098186,
+      "learning_rate": 4.202395402880921e-06,
+      "loss": 0.477,
+      "step": 8861
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7166849866313526,
+      "learning_rate": 4.201377043578103e-06,
+      "loss": 0.4841,
+      "step": 8862
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5869151770572896,
+      "learning_rate": 4.200358718270114e-06,
+      "loss": 0.4616,
+      "step": 8863
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7906584046026495,
+      "learning_rate": 4.199340427000298e-06,
+      "loss": 0.5174,
+      "step": 8864
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7584836030933219,
+      "learning_rate": 4.198322169812002e-06,
+      "loss": 0.5124,
+      "step": 8865
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7853310210595729,
+      "learning_rate": 4.197303946748569e-06,
+      "loss": 0.4599,
+      "step": 8866
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.677042962099927,
+      "learning_rate": 4.196285757853342e-06,
+      "loss": 0.4626,
+      "step": 8867
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.589929344842882,
+      "learning_rate": 4.195267603169662e-06,
+      "loss": 0.4007,
+      "step": 8868
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5947996911596473,
+      "learning_rate": 4.1942494827408685e-06,
+      "loss": 0.4221,
+      "step": 8869
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6595734885032779,
+      "learning_rate": 4.1932313966103e-06,
+      "loss": 0.4976,
+      "step": 8870
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.1396244402972773,
+      "learning_rate": 4.192213344821293e-06,
+      "loss": 0.4839,
+      "step": 8871
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.588643884617622,
+      "learning_rate": 4.191195327417179e-06,
+      "loss": 0.4778,
+      "step": 8872
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5849963640387403,
+      "learning_rate": 4.190177344441299e-06,
+      "loss": 0.4616,
+      "step": 8873
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6118014658680612,
+      "learning_rate": 4.189159395936982e-06,
+      "loss": 0.4733,
+      "step": 8874
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5344916359361866,
+      "learning_rate": 4.188141481947558e-06,
+      "loss": 0.409,
+      "step": 8875
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5913422283984344,
+      "learning_rate": 4.1871236025163555e-06,
+      "loss": 0.503,
+      "step": 8876
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7905925305725864,
+      "learning_rate": 4.186105757686705e-06,
+      "loss": 0.5289,
+      "step": 8877
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6753084030534934,
+      "learning_rate": 4.185087947501931e-06,
+      "loss": 0.4665,
+      "step": 8878
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5571779321620085,
+      "learning_rate": 4.184070172005359e-06,
+      "loss": 0.4388,
+      "step": 8879
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6044655266484702,
+      "learning_rate": 4.183052431240312e-06,
+      "loss": 0.4475,
+      "step": 8880
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.2723111294082243,
+      "learning_rate": 4.182034725250114e-06,
+      "loss": 0.4747,
+      "step": 8881
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.0176464400152592,
+      "learning_rate": 4.1810170540780826e-06,
+      "loss": 0.4841,
+      "step": 8882
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7675081835243758,
+      "learning_rate": 4.179999417767539e-06,
+      "loss": 0.5361,
+      "step": 8883
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7027056699374664,
+      "learning_rate": 4.178981816361799e-06,
+      "loss": 0.4888,
+      "step": 8884
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6187308169977014,
+      "learning_rate": 4.177964249904179e-06,
+      "loss": 0.4736,
+      "step": 8885
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6993935695545946,
+      "learning_rate": 4.176946718437994e-06,
+      "loss": 0.5236,
+      "step": 8886
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8617606858911112,
+      "learning_rate": 4.175929222006558e-06,
+      "loss": 0.5814,
+      "step": 8887
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6173512784789451,
+      "learning_rate": 4.174911760653182e-06,
+      "loss": 0.4179,
+      "step": 8888
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6606884887732691,
+      "learning_rate": 4.1738943344211735e-06,
+      "loss": 0.4781,
+      "step": 8889
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.9421626490039012,
+      "learning_rate": 4.172876943353845e-06,
+      "loss": 0.5081,
+      "step": 8890
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6840009487430139,
+      "learning_rate": 4.171859587494502e-06,
+      "loss": 0.4728,
+      "step": 8891
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6360795948164695,
+      "learning_rate": 4.17084226688645e-06,
+      "loss": 0.4867,
+      "step": 8892
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8072297640581896,
+      "learning_rate": 4.169824981572993e-06,
+      "loss": 0.5327,
+      "step": 8893
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6302079145475248,
+      "learning_rate": 4.168807731597434e-06,
+      "loss": 0.4216,
+      "step": 8894
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6265915277445072,
+      "learning_rate": 4.167790517003074e-06,
+      "loss": 0.4466,
+      "step": 8895
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8306004570823654,
+      "learning_rate": 4.166773337833212e-06,
+      "loss": 0.5239,
+      "step": 8896
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.0538530611806352,
+      "learning_rate": 4.165756194131145e-06,
+      "loss": 0.51,
+      "step": 8897
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7418913940773039,
+      "learning_rate": 4.164739085940172e-06,
+      "loss": 0.5105,
+      "step": 8898
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6867392525629221,
+      "learning_rate": 4.163722013303587e-06,
+      "loss": 0.5311,
+      "step": 8899
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.738323713708081,
+      "learning_rate": 4.1627049762646845e-06,
+      "loss": 0.5327,
+      "step": 8900
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6522210674657647,
+      "learning_rate": 4.1616879748667545e-06,
+      "loss": 0.494,
+      "step": 8901
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8231186804607882,
+      "learning_rate": 4.1606710091530885e-06,
+      "loss": 0.4958,
+      "step": 8902
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5755750166470923,
+      "learning_rate": 4.159654079166976e-06,
+      "loss": 0.4581,
+      "step": 8903
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7916153775811734,
+      "learning_rate": 4.158637184951704e-06,
+      "loss": 0.5453,
+      "step": 8904
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8241823371799982,
+      "learning_rate": 4.157620326550558e-06,
+      "loss": 0.575,
+      "step": 8905
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8173129061707199,
+      "learning_rate": 4.156603504006824e-06,
+      "loss": 0.5308,
+      "step": 8906
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5174368334878502,
+      "learning_rate": 4.1555867173637825e-06,
+      "loss": 0.4307,
+      "step": 8907
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6941899279197069,
+      "learning_rate": 4.154569966664716e-06,
+      "loss": 0.5053,
+      "step": 8908
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7752276093695819,
+      "learning_rate": 4.1535532519529056e-06,
+      "loss": 0.4935,
+      "step": 8909
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5709264687011458,
+      "learning_rate": 4.152536573271625e-06,
+      "loss": 0.4682,
+      "step": 8910
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8372322005711126,
+      "learning_rate": 4.151519930664158e-06,
+      "loss": 0.5747,
+      "step": 8911
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.623406384653625,
+      "learning_rate": 4.150503324173775e-06,
+      "loss": 0.4411,
+      "step": 8912
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6222006459899796,
+      "learning_rate": 4.14948675384375e-06,
+      "loss": 0.4718,
+      "step": 8913
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6049654092042297,
+      "learning_rate": 4.148470219717358e-06,
+      "loss": 0.4993,
+      "step": 8914
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7745050340158272,
+      "learning_rate": 4.147453721837866e-06,
+      "loss": 0.5273,
+      "step": 8915
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6109987895486814,
+      "learning_rate": 4.146437260248546e-06,
+      "loss": 0.4503,
+      "step": 8916
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6793429391527532,
+      "learning_rate": 4.145420834992664e-06,
+      "loss": 0.5184,
+      "step": 8917
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8818497858691048,
+      "learning_rate": 4.144404446113486e-06,
+      "loss": 0.5765,
+      "step": 8918
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.021306485742972,
+      "learning_rate": 4.143388093654276e-06,
+      "loss": 0.5162,
+      "step": 8919
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5859749858002101,
+      "learning_rate": 4.142371777658299e-06,
+      "loss": 0.4239,
+      "step": 8920
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6299049601946107,
+      "learning_rate": 4.141355498168813e-06,
+      "loss": 0.4581,
+      "step": 8921
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8455046076539325,
+      "learning_rate": 4.140339255229079e-06,
+      "loss": 0.5296,
+      "step": 8922
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.136195459596776,
+      "learning_rate": 4.139323048882357e-06,
+      "loss": 0.5579,
+      "step": 8923
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7686804183218642,
+      "learning_rate": 4.138306879171903e-06,
+      "loss": 0.5021,
+      "step": 8924
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6137399353408208,
+      "learning_rate": 4.137290746140972e-06,
+      "loss": 0.4575,
+      "step": 8925
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8048452011345688,
+      "learning_rate": 4.136274649832816e-06,
+      "loss": 0.4721,
+      "step": 8926
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6495430245986274,
+      "learning_rate": 4.135258590290688e-06,
+      "loss": 0.4444,
+      "step": 8927
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.847925245009506,
+      "learning_rate": 4.13424256755784e-06,
+      "loss": 0.5336,
+      "step": 8928
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6373667724772821,
+      "learning_rate": 4.133226581677518e-06,
+      "loss": 0.4617,
+      "step": 8929
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6544387934433573,
+      "learning_rate": 4.132210632692971e-06,
+      "loss": 0.4975,
+      "step": 8930
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.952005884589329,
+      "learning_rate": 4.131194720647445e-06,
+      "loss": 0.5441,
+      "step": 8931
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7160581309498305,
+      "learning_rate": 4.130178845584183e-06,
+      "loss": 0.4449,
+      "step": 8932
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.606984407071753,
+      "learning_rate": 4.129163007546427e-06,
+      "loss": 0.4834,
+      "step": 8933
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6049660768632178,
+      "learning_rate": 4.128147206577417e-06,
+      "loss": 0.4182,
+      "step": 8934
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.016503241966969,
+      "learning_rate": 4.127131442720398e-06,
+      "loss": 0.4922,
+      "step": 8935
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5893463205746421,
+      "learning_rate": 4.126115716018603e-06,
+      "loss": 0.4684,
+      "step": 8936
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6885041023236379,
+      "learning_rate": 4.12510002651527e-06,
+      "loss": 0.4548,
+      "step": 8937
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6365596791560154,
+      "learning_rate": 4.124084374253633e-06,
+      "loss": 0.4616,
+      "step": 8938
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7166848715376369,
+      "learning_rate": 4.123068759276925e-06,
+      "loss": 0.5603,
+      "step": 8939
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8134642852047768,
+      "learning_rate": 4.1220531816283775e-06,
+      "loss": 0.5208,
+      "step": 8940
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9409062362021231,
+      "learning_rate": 4.121037641351221e-06,
+      "loss": 0.5247,
+      "step": 8941
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5593972316517262,
+      "learning_rate": 4.120022138488683e-06,
+      "loss": 0.4282,
+      "step": 8942
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6206467823627109,
+      "learning_rate": 4.119006673083991e-06,
+      "loss": 0.5663,
+      "step": 8943
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.049580058284063,
+      "learning_rate": 4.117991245180369e-06,
+      "loss": 0.4577,
+      "step": 8944
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5608642755061072,
+      "learning_rate": 4.116975854821041e-06,
+      "loss": 0.4343,
+      "step": 8945
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6027099428056472,
+      "learning_rate": 4.11596050204923e-06,
+      "loss": 0.4791,
+      "step": 8946
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5888289668248263,
+      "learning_rate": 4.114945186908154e-06,
+      "loss": 0.4461,
+      "step": 8947
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5931289785112412,
+      "learning_rate": 4.113929909441034e-06,
+      "loss": 0.4489,
+      "step": 8948
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6288183169782241,
+      "learning_rate": 4.112914669691086e-06,
+      "loss": 0.466,
+      "step": 8949
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.130370526799863,
+      "learning_rate": 4.1118994677015265e-06,
+      "loss": 0.5343,
+      "step": 8950
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6846604108310098,
+      "learning_rate": 4.110884303515568e-06,
+      "loss": 0.4754,
+      "step": 8951
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.571214086548935,
+      "learning_rate": 4.109869177176424e-06,
+      "loss": 0.4288,
+      "step": 8952
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6284123393722311,
+      "learning_rate": 4.108854088727304e-06,
+      "loss": 0.4621,
+      "step": 8953
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.597164421679138,
+      "learning_rate": 4.107839038211419e-06,
+      "loss": 0.5202,
+      "step": 8954
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7273766474979937,
+      "learning_rate": 4.1068240256719746e-06,
+      "loss": 0.4573,
+      "step": 8955
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6700952731868057,
+      "learning_rate": 4.105809051152177e-06,
+      "loss": 0.4877,
+      "step": 8956
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9662626445861817,
+      "learning_rate": 4.104794114695231e-06,
+      "loss": 0.5201,
+      "step": 8957
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6337353218064731,
+      "learning_rate": 4.103779216344339e-06,
+      "loss": 0.4806,
+      "step": 8958
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.597347468555482,
+      "learning_rate": 4.1027643561427e-06,
+      "loss": 0.4529,
+      "step": 8959
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7032296208176871,
+      "learning_rate": 4.101749534133517e-06,
+      "loss": 0.5353,
+      "step": 8960
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8109327372478944,
+      "learning_rate": 4.100734750359986e-06,
+      "loss": 0.4953,
+      "step": 8961
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5384636345651621,
+      "learning_rate": 4.099720004865303e-06,
+      "loss": 0.4236,
+      "step": 8962
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5847172012755929,
+      "learning_rate": 4.098705297692662e-06,
+      "loss": 0.4492,
+      "step": 8963
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6224556095872419,
+      "learning_rate": 4.097690628885257e-06,
+      "loss": 0.4971,
+      "step": 8964
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8287472555661987,
+      "learning_rate": 4.096675998486278e-06,
+      "loss": 0.5301,
+      "step": 8965
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7877916671790234,
+      "learning_rate": 4.095661406538916e-06,
+      "loss": 0.5595,
+      "step": 8966
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6907176868516108,
+      "learning_rate": 4.094646853086357e-06,
+      "loss": 0.5248,
+      "step": 8967
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6387418884297833,
+      "learning_rate": 4.093632338171789e-06,
+      "loss": 0.5222,
+      "step": 8968
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6841759868077028,
+      "learning_rate": 4.092617861838396e-06,
+      "loss": 0.4741,
+      "step": 8969
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5932066233094653,
+      "learning_rate": 4.09160342412936e-06,
+      "loss": 0.4517,
+      "step": 8970
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5560793362032284,
+      "learning_rate": 4.090589025087864e-06,
+      "loss": 0.4687,
+      "step": 8971
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.67358657501909,
+      "learning_rate": 4.089574664757086e-06,
+      "loss": 0.503,
+      "step": 8972
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5601399046352287,
+      "learning_rate": 4.088560343180207e-06,
+      "loss": 0.4659,
+      "step": 8973
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5806752798821936,
+      "learning_rate": 4.087546060400401e-06,
+      "loss": 0.4736,
+      "step": 8974
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.890918080455701,
+      "learning_rate": 4.0865318164608435e-06,
+      "loss": 0.432,
+      "step": 8975
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6062355402362681,
+      "learning_rate": 4.085517611404708e-06,
+      "loss": 0.5202,
+      "step": 8976
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6896890431382324,
+      "learning_rate": 4.0845034452751656e-06,
+      "loss": 0.4868,
+      "step": 8977
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6243644930165145,
+      "learning_rate": 4.083489318115385e-06,
+      "loss": 0.4485,
+      "step": 8978
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6919195789594957,
+      "learning_rate": 4.082475229968536e-06,
+      "loss": 0.5431,
+      "step": 8979
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9770142347219757,
+      "learning_rate": 4.0814611808777845e-06,
+      "loss": 0.5338,
+      "step": 8980
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7213578510869765,
+      "learning_rate": 4.080447170886296e-06,
+      "loss": 0.481,
+      "step": 8981
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6027346606298167,
+      "learning_rate": 4.079433200037233e-06,
+      "loss": 0.4404,
+      "step": 8982
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6706439239375428,
+      "learning_rate": 4.078419268373756e-06,
+      "loss": 0.493,
+      "step": 8983
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.4632090335888754,
+      "learning_rate": 4.077405375939023e-06,
+      "loss": 0.5167,
+      "step": 8984
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5743036022268803,
+      "learning_rate": 4.076391522776198e-06,
+      "loss": 0.4724,
+      "step": 8985
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.000783417638258,
+      "learning_rate": 4.0753777089284345e-06,
+      "loss": 0.4938,
+      "step": 8986
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6789018943041462,
+      "learning_rate": 4.074363934438888e-06,
+      "loss": 0.4857,
+      "step": 8987
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.0172097982351374,
+      "learning_rate": 4.07335019935071e-06,
+      "loss": 0.503,
+      "step": 8988
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5806497672232582,
+      "learning_rate": 4.072336503707053e-06,
+      "loss": 0.4173,
+      "step": 8989
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6914722764442861,
+      "learning_rate": 4.071322847551066e-06,
+      "loss": 0.4729,
+      "step": 8990
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7195830021633124,
+      "learning_rate": 4.070309230925898e-06,
+      "loss": 0.4907,
+      "step": 8991
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6125785274309221,
+      "learning_rate": 4.0692956538746966e-06,
+      "loss": 0.4695,
+      "step": 8992
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6989050923253964,
+      "learning_rate": 4.068282116440604e-06,
+      "loss": 0.4747,
+      "step": 8993
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5832176441382122,
+      "learning_rate": 4.067268618666764e-06,
+      "loss": 0.4553,
+      "step": 8994
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6889027638096344,
+      "learning_rate": 4.066255160596319e-06,
+      "loss": 0.5304,
+      "step": 8995
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.761549075424479,
+      "learning_rate": 4.0652417422724054e-06,
+      "loss": 0.5378,
+      "step": 8996
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6770589232801114,
+      "learning_rate": 4.064228363738166e-06,
+      "loss": 0.5063,
+      "step": 8997
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6509939339017758,
+      "learning_rate": 4.063215025036735e-06,
+      "loss": 0.492,
+      "step": 8998
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8168004636442611,
+      "learning_rate": 4.062201726211247e-06,
+      "loss": 0.5236,
+      "step": 8999
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8033615596745731,
+      "learning_rate": 4.061188467304834e-06,
+      "loss": 0.5269,
+      "step": 9000
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.638583581326899,
+      "learning_rate": 4.060175248360629e-06,
+      "loss": 0.4864,
+      "step": 9001
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7500171028160009,
+      "learning_rate": 4.05916206942176e-06,
+      "loss": 0.4886,
+      "step": 9002
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6828589949118525,
+      "learning_rate": 4.058148930531355e-06,
+      "loss": 0.5003,
+      "step": 9003
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7868472436310205,
+      "learning_rate": 4.05713583173254e-06,
+      "loss": 0.5041,
+      "step": 9004
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6050353891825929,
+      "learning_rate": 4.056122773068441e-06,
+      "loss": 0.4544,
+      "step": 9005
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8083388533830902,
+      "learning_rate": 4.055109754582178e-06,
+      "loss": 0.4887,
+      "step": 9006
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.646573712630938,
+      "learning_rate": 4.054096776316874e-06,
+      "loss": 0.4645,
+      "step": 9007
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.672270820341946,
+      "learning_rate": 4.053083838315647e-06,
+      "loss": 0.4538,
+      "step": 9008
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6628567131546937,
+      "learning_rate": 4.052070940621614e-06,
+      "loss": 0.432,
+      "step": 9009
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6478913514779144,
+      "learning_rate": 4.051058083277894e-06,
+      "loss": 0.4497,
+      "step": 9010
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5845577247889107,
+      "learning_rate": 4.050045266327598e-06,
+      "loss": 0.4767,
+      "step": 9011
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.0573059363922057,
+      "learning_rate": 4.04903248981384e-06,
+      "loss": 0.5476,
+      "step": 9012
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6472651449472857,
+      "learning_rate": 4.0480197537797286e-06,
+      "loss": 0.4536,
+      "step": 9013
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7598329002180562,
+      "learning_rate": 4.047007058268376e-06,
+      "loss": 0.5106,
+      "step": 9014
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7673200262458939,
+      "learning_rate": 4.045994403322887e-06,
+      "loss": 0.5061,
+      "step": 9015
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.9184684277275242,
+      "learning_rate": 4.044981788986367e-06,
+      "loss": 0.4862,
+      "step": 9016
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6969550444167093,
+      "learning_rate": 4.043969215301922e-06,
+      "loss": 0.5409,
+      "step": 9017
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6221736352650029,
+      "learning_rate": 4.042956682312651e-06,
+      "loss": 0.476,
+      "step": 9018
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6134582174370081,
+      "learning_rate": 4.041944190061656e-06,
+      "loss": 0.4461,
+      "step": 9019
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.631557806281466,
+      "learning_rate": 4.040931738592036e-06,
+      "loss": 0.4557,
+      "step": 9020
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6065319619591251,
+      "learning_rate": 4.039919327946886e-06,
+      "loss": 0.4687,
+      "step": 9021
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7023971073960384,
+      "learning_rate": 4.038906958169303e-06,
+      "loss": 0.4781,
+      "step": 9022
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6913650205082353,
+      "learning_rate": 4.0378946293023796e-06,
+      "loss": 0.5183,
+      "step": 9023
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7805906534382324,
+      "learning_rate": 4.036882341389207e-06,
+      "loss": 0.443,
+      "step": 9024
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6412853220892517,
+      "learning_rate": 4.035870094472876e-06,
+      "loss": 0.4661,
+      "step": 9025
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5563209632543178,
+      "learning_rate": 4.034857888596474e-06,
+      "loss": 0.3932,
+      "step": 9026
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6111839346361122,
+      "learning_rate": 4.033845723803088e-06,
+      "loss": 0.4411,
+      "step": 9027
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6244846543392127,
+      "learning_rate": 4.032833600135802e-06,
+      "loss": 0.462,
+      "step": 9028
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6177985738441802,
+      "learning_rate": 4.031821517637698e-06,
+      "loss": 0.4518,
+      "step": 9029
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6938530739007701,
+      "learning_rate": 4.030809476351859e-06,
+      "loss": 0.4707,
+      "step": 9030
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6054436275252574,
+      "learning_rate": 4.029797476321363e-06,
+      "loss": 0.4833,
+      "step": 9031
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6034519654419125,
+      "learning_rate": 4.028785517589289e-06,
+      "loss": 0.4635,
+      "step": 9032
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8155015179816563,
+      "learning_rate": 4.027773600198711e-06,
+      "loss": 0.4319,
+      "step": 9033
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6682705107658354,
+      "learning_rate": 4.026761724192702e-06,
+      "loss": 0.4966,
+      "step": 9034
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7560401763168254,
+      "learning_rate": 4.02574988961434e-06,
+      "loss": 0.5027,
+      "step": 9035
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6071252331146583,
+      "learning_rate": 4.02473809650669e-06,
+      "loss": 0.4222,
+      "step": 9036
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5665434685751473,
+      "learning_rate": 4.023726344912824e-06,
+      "loss": 0.447,
+      "step": 9037
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5770185103714324,
+      "learning_rate": 4.0227146348758075e-06,
+      "loss": 0.5072,
+      "step": 9038
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5667922612344445,
+      "learning_rate": 4.021702966438705e-06,
+      "loss": 0.4384,
+      "step": 9039
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6097516268978369,
+      "learning_rate": 4.020691339644582e-06,
+      "loss": 0.492,
+      "step": 9040
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7460488808352627,
+      "learning_rate": 4.0196797545364995e-06,
+      "loss": 0.5028,
+      "step": 9041
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6187724371519613,
+      "learning_rate": 4.018668211157516e-06,
+      "loss": 0.4718,
+      "step": 9042
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7455951878050731,
+      "learning_rate": 4.017656709550693e-06,
+      "loss": 0.5081,
+      "step": 9043
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7889960882207994,
+      "learning_rate": 4.016645249759082e-06,
+      "loss": 0.4602,
+      "step": 9044
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6143357283415994,
+      "learning_rate": 4.0156338318257425e-06,
+      "loss": 0.4307,
+      "step": 9045
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5838916648152049,
+      "learning_rate": 4.014622455793723e-06,
+      "loss": 0.4506,
+      "step": 9046
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5911372939789319,
+      "learning_rate": 4.013611121706077e-06,
+      "loss": 0.5167,
+      "step": 9047
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8257912464304906,
+      "learning_rate": 4.012599829605855e-06,
+      "loss": 0.4686,
+      "step": 9048
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7150846440018089,
+      "learning_rate": 4.011588579536103e-06,
+      "loss": 0.5238,
+      "step": 9049
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8119351417033279,
+      "learning_rate": 4.010577371539867e-06,
+      "loss": 0.4893,
+      "step": 9050
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.626663596403873,
+      "learning_rate": 4.009566205660189e-06,
+      "loss": 0.4948,
+      "step": 9051
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8603723504016599,
+      "learning_rate": 4.0085550819401134e-06,
+      "loss": 0.5311,
+      "step": 9052
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.9767395437781328,
+      "learning_rate": 4.00754400042268e-06,
+      "loss": 0.5491,
+      "step": 9053
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8276632949212203,
+      "learning_rate": 4.006532961150927e-06,
+      "loss": 0.4969,
+      "step": 9054
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6460748235757424,
+      "learning_rate": 4.005521964167891e-06,
+      "loss": 0.4752,
+      "step": 9055
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6507742802554809,
+      "learning_rate": 4.004511009516607e-06,
+      "loss": 0.4548,
+      "step": 9056
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5536586439501516,
+      "learning_rate": 4.003500097240109e-06,
+      "loss": 0.4739,
+      "step": 9057
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6861642040176475,
+      "learning_rate": 4.002489227381425e-06,
+      "loss": 0.4602,
+      "step": 9058
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7315041685257905,
+      "learning_rate": 4.001478399983589e-06,
+      "loss": 0.4793,
+      "step": 9059
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5938745482269892,
+      "learning_rate": 4.000467615089626e-06,
+      "loss": 0.4507,
+      "step": 9060
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6054268412964247,
+      "learning_rate": 3.999456872742564e-06,
+      "loss": 0.4614,
+      "step": 9061
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8533290144798112,
+      "learning_rate": 3.9984461729854245e-06,
+      "loss": 0.4816,
+      "step": 9062
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7309372142370067,
+      "learning_rate": 3.997435515861231e-06,
+      "loss": 0.5166,
+      "step": 9063
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5973625572814367,
+      "learning_rate": 3.996424901413004e-06,
+      "loss": 0.4724,
+      "step": 9064
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7438344152897951,
+      "learning_rate": 3.995414329683762e-06,
+      "loss": 0.461,
+      "step": 9065
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5631106322421988,
+      "learning_rate": 3.994403800716523e-06,
+      "loss": 0.4277,
+      "step": 9066
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5838279354196818,
+      "learning_rate": 3.993393314554299e-06,
+      "loss": 0.4752,
+      "step": 9067
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7605875248332076,
+      "learning_rate": 3.992382871240106e-06,
+      "loss": 0.4666,
+      "step": 9068
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.680995340606293,
+      "learning_rate": 3.991372470816954e-06,
+      "loss": 0.4718,
+      "step": 9069
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5152597336678216,
+      "learning_rate": 3.990362113327853e-06,
+      "loss": 0.3887,
+      "step": 9070
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6079573837752223,
+      "learning_rate": 3.989351798815808e-06,
+      "loss": 0.4939,
+      "step": 9071
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8132652654058701,
+      "learning_rate": 3.98834152732383e-06,
+      "loss": 0.4989,
+      "step": 9072
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5725514724098311,
+      "learning_rate": 3.98733129889492e-06,
+      "loss": 0.4218,
+      "step": 9073
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5594307124063322,
+      "learning_rate": 3.98632111357208e-06,
+      "loss": 0.465,
+      "step": 9074
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5908550158977094,
+      "learning_rate": 3.98531097139831e-06,
+      "loss": 0.4296,
+      "step": 9075
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7342108400684618,
+      "learning_rate": 3.984300872416612e-06,
+      "loss": 0.552,
+      "step": 9076
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8040761807166786,
+      "learning_rate": 3.983290816669979e-06,
+      "loss": 0.5334,
+      "step": 9077
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7062037775856744,
+      "learning_rate": 3.982280804201407e-06,
+      "loss": 0.4834,
+      "step": 9078
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7694091459615919,
+      "learning_rate": 3.981270835053889e-06,
+      "loss": 0.474,
+      "step": 9079
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6418411681188796,
+      "learning_rate": 3.980260909270416e-06,
+      "loss": 0.4525,
+      "step": 9080
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5865001399814008,
+      "learning_rate": 3.979251026893977e-06,
+      "loss": 0.4726,
+      "step": 9081
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7546366528818873,
+      "learning_rate": 3.978241187967561e-06,
+      "loss": 0.5606,
+      "step": 9082
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7466575105866234,
+      "learning_rate": 3.977231392534152e-06,
+      "loss": 0.5011,
+      "step": 9083
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5675433021653249,
+      "learning_rate": 3.976221640636734e-06,
+      "loss": 0.4636,
+      "step": 9084
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6245798588680239,
+      "learning_rate": 3.9752119323182906e-06,
+      "loss": 0.4712,
+      "step": 9085
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8089754289607965,
+      "learning_rate": 3.974202267621799e-06,
+      "loss": 0.5458,
+      "step": 9086
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7455968135757399,
+      "learning_rate": 3.973192646590239e-06,
+      "loss": 0.5792,
+      "step": 9087
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.80761384602154,
+      "learning_rate": 3.972183069266588e-06,
+      "loss": 0.5336,
+      "step": 9088
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7661715973771724,
+      "learning_rate": 3.971173535693819e-06,
+      "loss": 0.4915,
+      "step": 9089
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6546664827639979,
+      "learning_rate": 3.9701640459149045e-06,
+      "loss": 0.471,
+      "step": 9090
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.736428828910854,
+      "learning_rate": 3.969154599972815e-06,
+      "loss": 0.449,
+      "step": 9091
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7050627946695277,
+      "learning_rate": 3.968145197910521e-06,
+      "loss": 0.5227,
+      "step": 9092
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8317858416370292,
+      "learning_rate": 3.967135839770989e-06,
+      "loss": 0.5211,
+      "step": 9093
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6304214899789228,
+      "learning_rate": 3.966126525597182e-06,
+      "loss": 0.4257,
+      "step": 9094
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5291623505855394,
+      "learning_rate": 3.965117255432067e-06,
+      "loss": 0.4246,
+      "step": 9095
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5870379339359486,
+      "learning_rate": 3.9641080293186005e-06,
+      "loss": 0.4106,
+      "step": 9096
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5793816523184112,
+      "learning_rate": 3.963098847299746e-06,
+      "loss": 0.4289,
+      "step": 9097
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.613706039667801,
+      "learning_rate": 3.962089709418463e-06,
+      "loss": 0.4932,
+      "step": 9098
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.963706234550816,
+      "learning_rate": 3.961080615717702e-06,
+      "loss": 0.4763,
+      "step": 9099
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7744957939857139,
+      "learning_rate": 3.96007156624042e-06,
+      "loss": 0.5497,
+      "step": 9100
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7003710811321566,
+      "learning_rate": 3.959062561029567e-06,
+      "loss": 0.5037,
+      "step": 9101
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6349705294957484,
+      "learning_rate": 3.9580536001280965e-06,
+      "loss": 0.4975,
+      "step": 9102
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6906702390940155,
+      "learning_rate": 3.957044683578953e-06,
+      "loss": 0.4872,
+      "step": 9103
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6341404581796939,
+      "learning_rate": 3.9560358114250855e-06,
+      "loss": 0.4845,
+      "step": 9104
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7150037390349223,
+      "learning_rate": 3.955026983709437e-06,
+      "loss": 0.4792,
+      "step": 9105
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7174431303440013,
+      "learning_rate": 3.954018200474951e-06,
+      "loss": 0.48,
+      "step": 9106
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7808515385455057,
+      "learning_rate": 3.953009461764568e-06,
+      "loss": 0.5819,
+      "step": 9107
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6942780320023652,
+      "learning_rate": 3.952000767621224e-06,
+      "loss": 0.4963,
+      "step": 9108
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.657704003477227,
+      "learning_rate": 3.95099211808786e-06,
+      "loss": 0.478,
+      "step": 9109
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7712681903048891,
+      "learning_rate": 3.94998351320741e-06,
+      "loss": 0.4958,
+      "step": 9110
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7358224316257641,
+      "learning_rate": 3.948974953022805e-06,
+      "loss": 0.5168,
+      "step": 9111
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6010775462509673,
+      "learning_rate": 3.94796643757698e-06,
+      "loss": 0.4673,
+      "step": 9112
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.819647934147269,
+      "learning_rate": 3.94695796691286e-06,
+      "loss": 0.5547,
+      "step": 9113
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7964932874161982,
+      "learning_rate": 3.945949541073376e-06,
+      "loss": 0.4886,
+      "step": 9114
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.602785597339483,
+      "learning_rate": 3.94494116010145e-06,
+      "loss": 0.4701,
+      "step": 9115
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.665245239858181,
+      "learning_rate": 3.943932824040009e-06,
+      "loss": 0.489,
+      "step": 9116
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.668834178179799,
+      "learning_rate": 3.942924532931971e-06,
+      "loss": 0.4795,
+      "step": 9117
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5957240798870477,
+      "learning_rate": 3.94191628682026e-06,
+      "loss": 0.4317,
+      "step": 9118
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5440271368903683,
+      "learning_rate": 3.94090808574779e-06,
+      "loss": 0.4569,
+      "step": 9119
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6572196855398421,
+      "learning_rate": 3.939899929757477e-06,
+      "loss": 0.4942,
+      "step": 9120
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9205188706946809,
+      "learning_rate": 3.938891818892238e-06,
+      "loss": 0.5233,
+      "step": 9121
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6757955028486089,
+      "learning_rate": 3.9378837531949834e-06,
+      "loss": 0.4662,
+      "step": 9122
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6469780203792238,
+      "learning_rate": 3.9368757327086235e-06,
+      "loss": 0.4821,
+      "step": 9123
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7221028375024128,
+      "learning_rate": 3.935867757476067e-06,
+      "loss": 0.5425,
+      "step": 9124
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8028123857059496,
+      "learning_rate": 3.934859827540219e-06,
+      "loss": 0.5813,
+      "step": 9125
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0441537245520633,
+      "learning_rate": 3.9338519429439844e-06,
+      "loss": 0.5078,
+      "step": 9126
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6024806031923067,
+      "learning_rate": 3.932844103730266e-06,
+      "loss": 0.4135,
+      "step": 9127
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5841228906941894,
+      "learning_rate": 3.931836309941964e-06,
+      "loss": 0.4907,
+      "step": 9128
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7006906670491828,
+      "learning_rate": 3.930828561621977e-06,
+      "loss": 0.509,
+      "step": 9129
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6420572105852815,
+      "learning_rate": 3.929820858813201e-06,
+      "loss": 0.4455,
+      "step": 9130
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6647095300568276,
+      "learning_rate": 3.928813201558531e-06,
+      "loss": 0.4785,
+      "step": 9131
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6692811303006562,
+      "learning_rate": 3.927805589900861e-06,
+      "loss": 0.5052,
+      "step": 9132
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7395503856659327,
+      "learning_rate": 3.926798023883079e-06,
+      "loss": 0.4945,
+      "step": 9133
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7046937754212744,
+      "learning_rate": 3.925790503548077e-06,
+      "loss": 0.5408,
+      "step": 9134
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7632893059144262,
+      "learning_rate": 3.92478302893874e-06,
+      "loss": 0.5102,
+      "step": 9135
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6120191728587682,
+      "learning_rate": 3.923775600097954e-06,
+      "loss": 0.4171,
+      "step": 9136
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6654767368305994,
+      "learning_rate": 3.922768217068601e-06,
+      "loss": 0.4044,
+      "step": 9137
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5990582930444863,
+      "learning_rate": 3.921760879893563e-06,
+      "loss": 0.4307,
+      "step": 9138
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6657303033587134,
+      "learning_rate": 3.92075358861572e-06,
+      "loss": 0.4622,
+      "step": 9139
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0007860509307496,
+      "learning_rate": 3.919746343277947e-06,
+      "loss": 0.5107,
+      "step": 9140
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5423788817410247,
+      "learning_rate": 3.91873914392312e-06,
+      "loss": 0.4407,
+      "step": 9141
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5877952194569003,
+      "learning_rate": 3.917731990594112e-06,
+      "loss": 0.4764,
+      "step": 9142
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8564332508298254,
+      "learning_rate": 3.916724883333796e-06,
+      "loss": 0.5384,
+      "step": 9143
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6409153378132896,
+      "learning_rate": 3.9157178221850395e-06,
+      "loss": 0.4464,
+      "step": 9144
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6126027797882431,
+      "learning_rate": 3.914710807190709e-06,
+      "loss": 0.4721,
+      "step": 9145
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6639521049575524,
+      "learning_rate": 3.913703838393673e-06,
+      "loss": 0.4303,
+      "step": 9146
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7090278906173495,
+      "learning_rate": 3.912696915836794e-06,
+      "loss": 0.4937,
+      "step": 9147
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.621386579863087,
+      "learning_rate": 3.911690039562931e-06,
+      "loss": 0.4277,
+      "step": 9148
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5783861492710738,
+      "learning_rate": 3.910683209614946e-06,
+      "loss": 0.4439,
+      "step": 9149
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6201388495073825,
+      "learning_rate": 3.909676426035696e-06,
+      "loss": 0.4661,
+      "step": 9150
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6921068944146899,
+      "learning_rate": 3.9086696888680365e-06,
+      "loss": 0.5297,
+      "step": 9151
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.697164699611184,
+      "learning_rate": 3.90766299815482e-06,
+      "loss": 0.4584,
+      "step": 9152
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8065126838874935,
+      "learning_rate": 3.9066563539389e-06,
+      "loss": 0.5355,
+      "step": 9153
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6212464792366286,
+      "learning_rate": 3.905649756263125e-06,
+      "loss": 0.4656,
+      "step": 9154
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5768977727785792,
+      "learning_rate": 3.904643205170343e-06,
+      "loss": 0.435,
+      "step": 9155
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6472479220189596,
+      "learning_rate": 3.903636700703399e-06,
+      "loss": 0.4406,
+      "step": 9156
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8617600728263312,
+      "learning_rate": 3.902630242905138e-06,
+      "loss": 0.5178,
+      "step": 9157
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7154616592155875,
+      "learning_rate": 3.901623831818398e-06,
+      "loss": 0.5068,
+      "step": 9158
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.766981964880095,
+      "learning_rate": 3.9006174674860245e-06,
+      "loss": 0.4771,
+      "step": 9159
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7299910472352982,
+      "learning_rate": 3.899611149950851e-06,
+      "loss": 0.4874,
+      "step": 9160
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6244080099514171,
+      "learning_rate": 3.898604879255715e-06,
+      "loss": 0.5004,
+      "step": 9161
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7549839214226711,
+      "learning_rate": 3.89759865544345e-06,
+      "loss": 0.5253,
+      "step": 9162
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6385097459167149,
+      "learning_rate": 3.896592478556888e-06,
+      "loss": 0.4846,
+      "step": 9163
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8405263514369358,
+      "learning_rate": 3.895586348638857e-06,
+      "loss": 0.4935,
+      "step": 9164
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7399865673635089,
+      "learning_rate": 3.894580265732187e-06,
+      "loss": 0.538,
+      "step": 9165
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6681394900829308,
+      "learning_rate": 3.893574229879702e-06,
+      "loss": 0.5062,
+      "step": 9166
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6105976858829424,
+      "learning_rate": 3.892568241124227e-06,
+      "loss": 0.4575,
+      "step": 9167
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5858101445608391,
+      "learning_rate": 3.891562299508582e-06,
+      "loss": 0.5049,
+      "step": 9168
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6234820883859475,
+      "learning_rate": 3.890556405075589e-06,
+      "loss": 0.4336,
+      "step": 9169
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7714491599082312,
+      "learning_rate": 3.889550557868062e-06,
+      "loss": 0.4954,
+      "step": 9170
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7779044214780643,
+      "learning_rate": 3.888544757928821e-06,
+      "loss": 0.5294,
+      "step": 9171
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7271296509389674,
+      "learning_rate": 3.887539005300679e-06,
+      "loss": 0.5461,
+      "step": 9172
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7822856196753434,
+      "learning_rate": 3.886533300026446e-06,
+      "loss": 0.573,
+      "step": 9173
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.733223695507951,
+      "learning_rate": 3.885527642148932e-06,
+      "loss": 0.4985,
+      "step": 9174
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7246966885432375,
+      "learning_rate": 3.884522031710946e-06,
+      "loss": 0.4717,
+      "step": 9175
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6986339145254724,
+      "learning_rate": 3.883516468755291e-06,
+      "loss": 0.5033,
+      "step": 9176
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6089393751958287,
+      "learning_rate": 3.882510953324773e-06,
+      "loss": 0.5221,
+      "step": 9177
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.3038668661326622,
+      "learning_rate": 3.881505485462192e-06,
+      "loss": 0.5175,
+      "step": 9178
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6827360566959162,
+      "learning_rate": 3.88050006521035e-06,
+      "loss": 0.5045,
+      "step": 9179
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5906429046719849,
+      "learning_rate": 3.879494692612041e-06,
+      "loss": 0.4377,
+      "step": 9180
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7615171443245319,
+      "learning_rate": 3.878489367710063e-06,
+      "loss": 0.5454,
+      "step": 9181
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7664159803596445,
+      "learning_rate": 3.877484090547207e-06,
+      "loss": 0.5827,
+      "step": 9182
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6761907687636795,
+      "learning_rate": 3.876478861166269e-06,
+      "loss": 0.5063,
+      "step": 9183
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7629195482273305,
+      "learning_rate": 3.875473679610034e-06,
+      "loss": 0.503,
+      "step": 9184
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6107465754141232,
+      "learning_rate": 3.874468545921292e-06,
+      "loss": 0.4959,
+      "step": 9185
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5943484056256512,
+      "learning_rate": 3.873463460142827e-06,
+      "loss": 0.476,
+      "step": 9186
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.456930780097185,
+      "learning_rate": 3.872458422317422e-06,
+      "loss": 0.5605,
+      "step": 9187
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6893922390539317,
+      "learning_rate": 3.871453432487859e-06,
+      "loss": 0.5372,
+      "step": 9188
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7219061276696814,
+      "learning_rate": 3.870448490696918e-06,
+      "loss": 0.5353,
+      "step": 9189
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6049954437753662,
+      "learning_rate": 3.869443596987374e-06,
+      "loss": 0.4549,
+      "step": 9190
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6344372661780289,
+      "learning_rate": 3.868438751402005e-06,
+      "loss": 0.5042,
+      "step": 9191
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6941526922751737,
+      "learning_rate": 3.867433953983582e-06,
+      "loss": 0.5014,
+      "step": 9192
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6881241446546059,
+      "learning_rate": 3.8664292047748755e-06,
+      "loss": 0.5136,
+      "step": 9193
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6162241616900622,
+      "learning_rate": 3.8654245038186556e-06,
+      "loss": 0.4564,
+      "step": 9194
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6233690103539943,
+      "learning_rate": 3.864419851157688e-06,
+      "loss": 0.4384,
+      "step": 9195
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.9418688162080785,
+      "learning_rate": 3.86341524683474e-06,
+      "loss": 0.5472,
+      "step": 9196
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8691137996594228,
+      "learning_rate": 3.862410690892572e-06,
+      "loss": 0.5088,
+      "step": 9197
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8002155652223986,
+      "learning_rate": 3.8614061833739465e-06,
+      "loss": 0.4869,
+      "step": 9198
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7042408303696801,
+      "learning_rate": 3.86040172432162e-06,
+      "loss": 0.5247,
+      "step": 9199
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5657596663045861,
+      "learning_rate": 3.859397313778352e-06,
+      "loss": 0.4336,
+      "step": 9200
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6596114493968095,
+      "learning_rate": 3.858392951786895e-06,
+      "loss": 0.4478,
+      "step": 9201
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6281678243658037,
+      "learning_rate": 3.857388638390001e-06,
+      "loss": 0.4427,
+      "step": 9202
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6567881099577151,
+      "learning_rate": 3.856384373630424e-06,
+      "loss": 0.471,
+      "step": 9203
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6881564327469285,
+      "learning_rate": 3.855380157550907e-06,
+      "loss": 0.495,
+      "step": 9204
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7740229586344956,
+      "learning_rate": 3.854375990194201e-06,
+      "loss": 0.5614,
+      "step": 9205
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7221591605665109,
+      "learning_rate": 3.853371871603048e-06,
+      "loss": 0.5433,
+      "step": 9206
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6764043248639546,
+      "learning_rate": 3.852367801820188e-06,
+      "loss": 0.519,
+      "step": 9207
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8623854662145037,
+      "learning_rate": 3.851363780888365e-06,
+      "loss": 0.5967,
+      "step": 9208
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7071161215088375,
+      "learning_rate": 3.8503598088503155e-06,
+      "loss": 0.5319,
+      "step": 9209
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7849210922790041,
+      "learning_rate": 3.849355885748775e-06,
+      "loss": 0.4994,
+      "step": 9210
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5831460860868768,
+      "learning_rate": 3.848352011626478e-06,
+      "loss": 0.4645,
+      "step": 9211
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5863534498201394,
+      "learning_rate": 3.847348186526156e-06,
+      "loss": 0.4421,
+      "step": 9212
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5815739507435796,
+      "learning_rate": 3.846344410490538e-06,
+      "loss": 0.4289,
+      "step": 9213
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5997127376286291,
+      "learning_rate": 3.845340683562352e-06,
+      "loss": 0.4683,
+      "step": 9214
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5790066090762571,
+      "learning_rate": 3.844337005784322e-06,
+      "loss": 0.432,
+      "step": 9215
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6239345958018846,
+      "learning_rate": 3.843333377199173e-06,
+      "loss": 0.4789,
+      "step": 9216
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8485273250638531,
+      "learning_rate": 3.8423297978496274e-06,
+      "loss": 0.5108,
+      "step": 9217
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5536674072142241,
+      "learning_rate": 3.841326267778403e-06,
+      "loss": 0.4159,
+      "step": 9218
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6455557634131391,
+      "learning_rate": 3.840322787028216e-06,
+      "loss": 0.4017,
+      "step": 9219
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.617889907250975,
+      "learning_rate": 3.83931935564178e-06,
+      "loss": 0.4648,
+      "step": 9220
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7682929414251215,
+      "learning_rate": 3.838315973661812e-06,
+      "loss": 0.5482,
+      "step": 9221
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7356949883681463,
+      "learning_rate": 3.83731264113102e-06,
+      "loss": 0.5363,
+      "step": 9222
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7679844367820574,
+      "learning_rate": 3.836309358092115e-06,
+      "loss": 0.5493,
+      "step": 9223
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8391703098611941,
+      "learning_rate": 3.835306124587801e-06,
+      "loss": 0.5955,
+      "step": 9224
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6816100992227841,
+      "learning_rate": 3.834302940660784e-06,
+      "loss": 0.5169,
+      "step": 9225
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6352445804077987,
+      "learning_rate": 3.8332998063537656e-06,
+      "loss": 0.4862,
+      "step": 9226
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7416147191657123,
+      "learning_rate": 3.832296721709447e-06,
+      "loss": 0.4999,
+      "step": 9227
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6579608237460424,
+      "learning_rate": 3.831293686770524e-06,
+      "loss": 0.4468,
+      "step": 9228
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6927788150118354,
+      "learning_rate": 3.830290701579695e-06,
+      "loss": 0.4558,
+      "step": 9229
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7305368181712655,
+      "learning_rate": 3.829287766179653e-06,
+      "loss": 0.4848,
+      "step": 9230
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.809619839555621,
+      "learning_rate": 3.82828488061309e-06,
+      "loss": 0.4916,
+      "step": 9231
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.2811635489446969,
+      "learning_rate": 3.827282044922692e-06,
+      "loss": 0.4731,
+      "step": 9232
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6003447374299552,
+      "learning_rate": 3.826279259151153e-06,
+      "loss": 0.4821,
+      "step": 9233
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7247723980490272,
+      "learning_rate": 3.8252765233411545e-06,
+      "loss": 0.445,
+      "step": 9234
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6432466770235389,
+      "learning_rate": 3.824273837535381e-06,
+      "loss": 0.4595,
+      "step": 9235
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8148596041799379,
+      "learning_rate": 3.8232712017765136e-06,
+      "loss": 0.569,
+      "step": 9236
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8379815591291666,
+      "learning_rate": 3.822268616107231e-06,
+      "loss": 0.4709,
+      "step": 9237
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6596493905406737,
+      "learning_rate": 3.821266080570209e-06,
+      "loss": 0.453,
+      "step": 9238
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7008662707900963,
+      "learning_rate": 3.8202635952081235e-06,
+      "loss": 0.4747,
+      "step": 9239
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6988174692851142,
+      "learning_rate": 3.8192611600636475e-06,
+      "loss": 0.4747,
+      "step": 9240
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7049989135272603,
+      "learning_rate": 3.81825877517945e-06,
+      "loss": 0.5187,
+      "step": 9241
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6617713091278571,
+      "learning_rate": 3.8172564405982e-06,
+      "loss": 0.5134,
+      "step": 9242
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6510452903445328,
+      "learning_rate": 3.816254156362565e-06,
+      "loss": 0.5014,
+      "step": 9243
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7906560317435627,
+      "learning_rate": 3.815251922515205e-06,
+      "loss": 0.5015,
+      "step": 9244
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8100131437901296,
+      "learning_rate": 3.814249739098787e-06,
+      "loss": 0.5174,
+      "step": 9245
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5886019392354486,
+      "learning_rate": 3.8132476061559683e-06,
+      "loss": 0.4444,
+      "step": 9246
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6982257396631043,
+      "learning_rate": 3.8122455237294065e-06,
+      "loss": 0.4943,
+      "step": 9247
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7774025952607653,
+      "learning_rate": 3.811243491861758e-06,
+      "loss": 0.5106,
+      "step": 9248
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8108360143504353,
+      "learning_rate": 3.8102415105956746e-06,
+      "loss": 0.5042,
+      "step": 9249
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7335364714964374,
+      "learning_rate": 3.8092395799738084e-06,
+      "loss": 0.5216,
+      "step": 9250
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7635895452982631,
+      "learning_rate": 3.8082377000388083e-06,
+      "loss": 0.447,
+      "step": 9251
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7662973614194442,
+      "learning_rate": 3.807235870833321e-06,
+      "loss": 0.5446,
+      "step": 9252
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7325856723178641,
+      "learning_rate": 3.8062340923999906e-06,
+      "loss": 0.5315,
+      "step": 9253
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7968720456675714,
+      "learning_rate": 3.8052323647814604e-06,
+      "loss": 0.5503,
+      "step": 9254
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7279809041346655,
+      "learning_rate": 3.8042306880203706e-06,
+      "loss": 0.4815,
+      "step": 9255
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6143506890845476,
+      "learning_rate": 3.803229062159359e-06,
+      "loss": 0.4539,
+      "step": 9256
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6321351488723453,
+      "learning_rate": 3.8022274872410614e-06,
+      "loss": 0.5123,
+      "step": 9257
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7688371835938946,
+      "learning_rate": 3.8012259633081133e-06,
+      "loss": 0.5241,
+      "step": 9258
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.655250172919114,
+      "learning_rate": 3.800224490403145e-06,
+      "loss": 0.4987,
+      "step": 9259
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7011900613464648,
+      "learning_rate": 3.7992230685687845e-06,
+      "loss": 0.5243,
+      "step": 9260
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7145118081784876,
+      "learning_rate": 3.798221697847663e-06,
+      "loss": 0.4575,
+      "step": 9261
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5398549303443795,
+      "learning_rate": 3.7972203782824023e-06,
+      "loss": 0.4333,
+      "step": 9262
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8049330864013907,
+      "learning_rate": 3.7962191099156277e-06,
+      "loss": 0.4919,
+      "step": 9263
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7253450753326646,
+      "learning_rate": 3.795217892789958e-06,
+      "loss": 0.5375,
+      "step": 9264
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7160461422226329,
+      "learning_rate": 3.7942167269480128e-06,
+      "loss": 0.5535,
+      "step": 9265
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7149180236707086,
+      "learning_rate": 3.793215612432409e-06,
+      "loss": 0.5485,
+      "step": 9266
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7221360989074198,
+      "learning_rate": 3.7922145492857597e-06,
+      "loss": 0.5037,
+      "step": 9267
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6515305820180698,
+      "learning_rate": 3.7912135375506774e-06,
+      "loss": 0.5286,
+      "step": 9268
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7626597267477976,
+      "learning_rate": 3.790212577269772e-06,
+      "loss": 0.5919,
+      "step": 9269
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6999679969840675,
+      "learning_rate": 3.789211668485651e-06,
+      "loss": 0.5081,
+      "step": 9270
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6184696543695805,
+      "learning_rate": 3.788210811240921e-06,
+      "loss": 0.4932,
+      "step": 9271
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8672299006600014,
+      "learning_rate": 3.787210005578184e-06,
+      "loss": 0.5358,
+      "step": 9272
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.1937749189019775,
+      "learning_rate": 3.7862092515400416e-06,
+      "loss": 0.5308,
+      "step": 9273
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7309519734392892,
+      "learning_rate": 3.7852085491690925e-06,
+      "loss": 0.5436,
+      "step": 9274
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.2516075414710495,
+      "learning_rate": 3.784207898507934e-06,
+      "loss": 0.4916,
+      "step": 9275
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6548281765815485,
+      "learning_rate": 3.7832072995991597e-06,
+      "loss": 0.4572,
+      "step": 9276
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6568133177718134,
+      "learning_rate": 3.782206752485363e-06,
+      "loss": 0.44,
+      "step": 9277
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.580457937503196,
+      "learning_rate": 3.781206257209133e-06,
+      "loss": 0.4221,
+      "step": 9278
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5952395096909264,
+      "learning_rate": 3.7802058138130577e-06,
+      "loss": 0.4786,
+      "step": 9279
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7074837466214744,
+      "learning_rate": 3.779205422339723e-06,
+      "loss": 0.4749,
+      "step": 9280
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7039139181821368,
+      "learning_rate": 3.7782050828317123e-06,
+      "loss": 0.471,
+      "step": 9281
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8467539821181148,
+      "learning_rate": 3.777204795331605e-06,
+      "loss": 0.5201,
+      "step": 9282
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6330137298086674,
+      "learning_rate": 3.776204559881984e-06,
+      "loss": 0.4785,
+      "step": 9283
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7968895662099147,
+      "learning_rate": 3.7752043765254248e-06,
+      "loss": 0.5121,
+      "step": 9284
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7022791635208689,
+      "learning_rate": 3.7742042453045014e-06,
+      "loss": 0.4924,
+      "step": 9285
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.779389833041283,
+      "learning_rate": 3.773204166261785e-06,
+      "loss": 0.4963,
+      "step": 9286
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6689910998117216,
+      "learning_rate": 3.772204139439848e-06,
+      "loss": 0.5007,
+      "step": 9287
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6497281707322332,
+      "learning_rate": 3.771204164881257e-06,
+      "loss": 0.5289,
+      "step": 9288
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7494177434850453,
+      "learning_rate": 3.7702042426285783e-06,
+      "loss": 0.5086,
+      "step": 9289
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5489349441227107,
+      "learning_rate": 3.769204372724375e-06,
+      "loss": 0.4177,
+      "step": 9290
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7714326818527153,
+      "learning_rate": 3.7682045552112084e-06,
+      "loss": 0.4482,
+      "step": 9291
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.1313778681595092,
+      "learning_rate": 3.7672047901316377e-06,
+      "loss": 0.5338,
+      "step": 9292
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7377633912869004,
+      "learning_rate": 3.766205077528219e-06,
+      "loss": 0.5001,
+      "step": 9293
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7411550229133614,
+      "learning_rate": 3.7652054174435065e-06,
+      "loss": 0.5159,
+      "step": 9294
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9335382124106479,
+      "learning_rate": 3.7642058099200553e-06,
+      "loss": 0.5174,
+      "step": 9295
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6597320538383452,
+      "learning_rate": 3.7632062550004135e-06,
+      "loss": 0.4275,
+      "step": 9296
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6055237739251296,
+      "learning_rate": 3.7622067527271288e-06,
+      "loss": 0.4492,
+      "step": 9297
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8273284377787128,
+      "learning_rate": 3.7612073031427484e-06,
+      "loss": 0.5092,
+      "step": 9298
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6844810617536631,
+      "learning_rate": 3.7602079062898133e-06,
+      "loss": 0.5355,
+      "step": 9299
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8157952071011444,
+      "learning_rate": 3.7592085622108666e-06,
+      "loss": 0.501,
+      "step": 9300
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5685104765345936,
+      "learning_rate": 3.758209270948446e-06,
+      "loss": 0.4738,
+      "step": 9301
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6627901912611,
+      "learning_rate": 3.7572100325450883e-06,
+      "loss": 0.5089,
+      "step": 9302
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7594495353184083,
+      "learning_rate": 3.756210847043329e-06,
+      "loss": 0.5367,
+      "step": 9303
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7573227987303697,
+      "learning_rate": 3.755211714485698e-06,
+      "loss": 0.5125,
+      "step": 9304
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8435916604362016,
+      "learning_rate": 3.7542126349147277e-06,
+      "loss": 0.5543,
+      "step": 9305
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5922881331706207,
+      "learning_rate": 3.753213608372942e-06,
+      "loss": 0.4347,
+      "step": 9306
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8723259026302791,
+      "learning_rate": 3.752214634902871e-06,
+      "loss": 0.4632,
+      "step": 9307
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.1746982021610626,
+      "learning_rate": 3.751215714547036e-06,
+      "loss": 0.5416,
+      "step": 9308
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7410483229901984,
+      "learning_rate": 3.7502168473479572e-06,
+      "loss": 0.4885,
+      "step": 9309
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6948772900516106,
+      "learning_rate": 3.749218033348153e-06,
+      "loss": 0.4579,
+      "step": 9310
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.9095252001287359,
+      "learning_rate": 3.748219272590141e-06,
+      "loss": 0.5086,
+      "step": 9311
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6682306627158009,
+      "learning_rate": 3.747220565116434e-06,
+      "loss": 0.4828,
+      "step": 9312
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6265989931863598,
+      "learning_rate": 3.7462219109695443e-06,
+      "loss": 0.4547,
+      "step": 9313
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6501863023514232,
+      "learning_rate": 3.7452233101919815e-06,
+      "loss": 0.5018,
+      "step": 9314
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6557278141796754,
+      "learning_rate": 3.744224762826253e-06,
+      "loss": 0.4884,
+      "step": 9315
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5893417255077114,
+      "learning_rate": 3.7432262689148625e-06,
+      "loss": 0.4312,
+      "step": 9316
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5455050742315564,
+      "learning_rate": 3.7422278285003145e-06,
+      "loss": 0.4005,
+      "step": 9317
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6535238794781585,
+      "learning_rate": 3.7412294416251084e-06,
+      "loss": 0.4468,
+      "step": 9318
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6805247278074009,
+      "learning_rate": 3.7402311083317413e-06,
+      "loss": 0.4741,
+      "step": 9319
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8638228911588726,
+      "learning_rate": 3.739232828662711e-06,
+      "loss": 0.563,
+      "step": 9320
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8193523934847567,
+      "learning_rate": 3.7382346026605108e-06,
+      "loss": 0.5108,
+      "step": 9321
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8829775620531921,
+      "learning_rate": 3.73723643036763e-06,
+      "loss": 0.4882,
+      "step": 9322
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5870229389234177,
+      "learning_rate": 3.736238311826561e-06,
+      "loss": 0.4887,
+      "step": 9323
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6344721944487332,
+      "learning_rate": 3.735240247079788e-06,
+      "loss": 0.505,
+      "step": 9324
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6926947312885636,
+      "learning_rate": 3.7342422361697967e-06,
+      "loss": 0.5369,
+      "step": 9325
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7269810014722693,
+      "learning_rate": 3.733244279139068e-06,
+      "loss": 0.4839,
+      "step": 9326
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.610876673986871,
+      "learning_rate": 3.732246376030083e-06,
+      "loss": 0.5129,
+      "step": 9327
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7641989057294315,
+      "learning_rate": 3.7312485268853184e-06,
+      "loss": 0.5836,
+      "step": 9328
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6870809711148789,
+      "learning_rate": 3.7302507317472493e-06,
+      "loss": 0.4601,
+      "step": 9329
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6202871951475808,
+      "learning_rate": 3.7292529906583487e-06,
+      "loss": 0.5205,
+      "step": 9330
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7142741815292583,
+      "learning_rate": 3.728255303661087e-06,
+      "loss": 0.5074,
+      "step": 9331
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5518207667321977,
+      "learning_rate": 3.7272576707979345e-06,
+      "loss": 0.4166,
+      "step": 9332
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6242109176807179,
+      "learning_rate": 3.7262600921113555e-06,
+      "loss": 0.5183,
+      "step": 9333
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7938759588234392,
+      "learning_rate": 3.7252625676438136e-06,
+      "loss": 0.5432,
+      "step": 9334
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7667275698528835,
+      "learning_rate": 3.7242650974377716e-06,
+      "loss": 0.5116,
+      "step": 9335
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.651678924316612,
+      "learning_rate": 3.7232676815356873e-06,
+      "loss": 0.4523,
+      "step": 9336
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8319730542316909,
+      "learning_rate": 3.722270319980018e-06,
+      "loss": 0.4918,
+      "step": 9337
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6840768630249385,
+      "learning_rate": 3.721273012813218e-06,
+      "loss": 0.525,
+      "step": 9338
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6739512175000334,
+      "learning_rate": 3.72027576007774e-06,
+      "loss": 0.4692,
+      "step": 9339
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6622551503471159,
+      "learning_rate": 3.7192785618160334e-06,
+      "loss": 0.4986,
+      "step": 9340
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7407529112361289,
+      "learning_rate": 3.7182814180705458e-06,
+      "loss": 0.5132,
+      "step": 9341
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6822790479885465,
+      "learning_rate": 3.7172843288837225e-06,
+      "loss": 0.4565,
+      "step": 9342
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6566713989972134,
+      "learning_rate": 3.7162872942980054e-06,
+      "loss": 0.4618,
+      "step": 9343
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6037750803989895,
+      "learning_rate": 3.7152903143558348e-06,
+      "loss": 0.4513,
+      "step": 9344
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7822450616837964,
+      "learning_rate": 3.7142933890996524e-06,
+      "loss": 0.4982,
+      "step": 9345
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.129460569369605,
+      "learning_rate": 3.7132965185718916e-06,
+      "loss": 0.5558,
+      "step": 9346
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.628171536754497,
+      "learning_rate": 3.7122997028149867e-06,
+      "loss": 0.4233,
+      "step": 9347
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6393485393375593,
+      "learning_rate": 3.7113029418713677e-06,
+      "loss": 0.4723,
+      "step": 9348
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5747234858157932,
+      "learning_rate": 3.710306235783465e-06,
+      "loss": 0.4561,
+      "step": 9349
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.59850375244888,
+      "learning_rate": 3.7093095845937043e-06,
+      "loss": 0.4571,
+      "step": 9350
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7406742826051419,
+      "learning_rate": 3.70831298834451e-06,
+      "loss": 0.4595,
+      "step": 9351
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8151994076648557,
+      "learning_rate": 3.7073164470783053e-06,
+      "loss": 0.49,
+      "step": 9352
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6471894018644406,
+      "learning_rate": 3.706319960837507e-06,
+      "loss": 0.4833,
+      "step": 9353
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7934390458550817,
+      "learning_rate": 3.705323529664535e-06,
+      "loss": 0.5053,
+      "step": 9354
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6241388647738333,
+      "learning_rate": 3.7043271536018033e-06,
+      "loss": 0.4895,
+      "step": 9355
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6094969019526522,
+      "learning_rate": 3.7033308326917216e-06,
+      "loss": 0.5041,
+      "step": 9356
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6697576398262278,
+      "learning_rate": 3.7023345669767047e-06,
+      "loss": 0.468,
+      "step": 9357
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5401047491143914,
+      "learning_rate": 3.701338356499159e-06,
+      "loss": 0.4228,
+      "step": 9358
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.583904784802937,
+      "learning_rate": 3.7003422013014887e-06,
+      "loss": 0.4863,
+      "step": 9359
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.9277342719980841,
+      "learning_rate": 3.699346101426099e-06,
+      "loss": 0.53,
+      "step": 9360
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7715251346295013,
+      "learning_rate": 3.698350056915389e-06,
+      "loss": 0.5672,
+      "step": 9361
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.4160447402707952,
+      "learning_rate": 3.6973540678117577e-06,
+      "loss": 0.5241,
+      "step": 9362
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5779429888948027,
+      "learning_rate": 3.6963581341576005e-06,
+      "loss": 0.4098,
+      "step": 9363
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6637529487488397,
+      "learning_rate": 3.695362255995312e-06,
+      "loss": 0.5328,
+      "step": 9364
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7283515188383624,
+      "learning_rate": 3.6943664333672835e-06,
+      "loss": 0.5162,
+      "step": 9365
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6114779598947504,
+      "learning_rate": 3.693370666315904e-06,
+      "loss": 0.4714,
+      "step": 9366
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6264892646274346,
+      "learning_rate": 3.692374954883559e-06,
+      "loss": 0.4412,
+      "step": 9367
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5449127386935234,
+      "learning_rate": 3.691379299112632e-06,
+      "loss": 0.4392,
+      "step": 9368
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6486514898573553,
+      "learning_rate": 3.6903836990455087e-06,
+      "loss": 0.449,
+      "step": 9369
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7254549508570198,
+      "learning_rate": 3.6893881547245658e-06,
+      "loss": 0.4963,
+      "step": 9370
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7918417743413897,
+      "learning_rate": 3.688392666192182e-06,
+      "loss": 0.4872,
+      "step": 9371
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5722716888119294,
+      "learning_rate": 3.6873972334907303e-06,
+      "loss": 0.4483,
+      "step": 9372
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6883569104658412,
+      "learning_rate": 3.686401856662584e-06,
+      "loss": 0.4523,
+      "step": 9373
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6141976950426694,
+      "learning_rate": 3.6854065357501133e-06,
+      "loss": 0.4703,
+      "step": 9374
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7885949112541654,
+      "learning_rate": 3.6844112707956852e-06,
+      "loss": 0.5138,
+      "step": 9375
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.5029589723466967,
+      "learning_rate": 3.683416061841665e-06,
+      "loss": 0.5571,
+      "step": 9376
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6642345677986007,
+      "learning_rate": 3.6824209089304163e-06,
+      "loss": 0.5051,
+      "step": 9377
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6578556018806611,
+      "learning_rate": 3.6814258121042995e-06,
+      "loss": 0.5162,
+      "step": 9378
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6995339114569948,
+      "learning_rate": 3.680430771405672e-06,
+      "loss": 0.4896,
+      "step": 9379
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5122522184510796,
+      "learning_rate": 3.6794357868768895e-06,
+      "loss": 0.4089,
+      "step": 9380
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5171705227254032,
+      "learning_rate": 3.6784408585603047e-06,
+      "loss": 0.4283,
+      "step": 9381
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.578098014850233,
+      "learning_rate": 3.677445986498271e-06,
+      "loss": 0.4353,
+      "step": 9382
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6636459750018662,
+      "learning_rate": 3.6764511707331354e-06,
+      "loss": 0.4648,
+      "step": 9383
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7143341456500879,
+      "learning_rate": 3.6754564113072423e-06,
+      "loss": 0.541,
+      "step": 9384
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.623908847572535,
+      "learning_rate": 3.6744617082629387e-06,
+      "loss": 0.4844,
+      "step": 9385
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7033864445488113,
+      "learning_rate": 3.6734670616425638e-06,
+      "loss": 0.551,
+      "step": 9386
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7838962325925771,
+      "learning_rate": 3.6724724714884574e-06,
+      "loss": 0.5106,
+      "step": 9387
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6528243369407437,
+      "learning_rate": 3.6714779378429556e-06,
+      "loss": 0.4701,
+      "step": 9388
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.659762743757758,
+      "learning_rate": 3.670483460748393e-06,
+      "loss": 0.4813,
+      "step": 9389
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7722851731699196,
+      "learning_rate": 3.6694890402471005e-06,
+      "loss": 0.5959,
+      "step": 9390
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8287668910847963,
+      "learning_rate": 3.6684946763814078e-06,
+      "loss": 0.5985,
+      "step": 9391
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7381844742000775,
+      "learning_rate": 3.6675003691936423e-06,
+      "loss": 0.5472,
+      "step": 9392
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7228108698755362,
+      "learning_rate": 3.666506118726127e-06,
+      "loss": 0.4872,
+      "step": 9393
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6189659019990906,
+      "learning_rate": 3.665511925021186e-06,
+      "loss": 0.5008,
+      "step": 9394
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6769923148269256,
+      "learning_rate": 3.6645177881211375e-06,
+      "loss": 0.476,
+      "step": 9395
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7445779339363285,
+      "learning_rate": 3.6635237080682994e-06,
+      "loss": 0.4593,
+      "step": 9396
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5909218732290121,
+      "learning_rate": 3.6625296849049863e-06,
+      "loss": 0.4433,
+      "step": 9397
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7741790992713794,
+      "learning_rate": 3.661535718673511e-06,
+      "loss": 0.5006,
+      "step": 9398
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6337812616324129,
+      "learning_rate": 3.660541809416182e-06,
+      "loss": 0.527,
+      "step": 9399
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8041060394041263,
+      "learning_rate": 3.659547957175309e-06,
+      "loss": 0.4815,
+      "step": 9400
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6972959503201985,
+      "learning_rate": 3.6585541619931953e-06,
+      "loss": 0.5485,
+      "step": 9401
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5640436797815905,
+      "learning_rate": 3.6575604239121444e-06,
+      "loss": 0.4168,
+      "step": 9402
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8431772662329683,
+      "learning_rate": 3.6565667429744565e-06,
+      "loss": 0.4642,
+      "step": 9403
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.787329924089397,
+      "learning_rate": 3.6555731192224294e-06,
+      "loss": 0.5255,
+      "step": 9404
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7807554087944258,
+      "learning_rate": 3.6545795526983584e-06,
+      "loss": 0.562,
+      "step": 9405
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5846642178958515,
+      "learning_rate": 3.6535860434445343e-06,
+      "loss": 0.462,
+      "step": 9406
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5956953735044148,
+      "learning_rate": 3.652592591503252e-06,
+      "loss": 0.5001,
+      "step": 9407
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6812276395809893,
+      "learning_rate": 3.6515991969167975e-06,
+      "loss": 0.4857,
+      "step": 9408
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6365016702927165,
+      "learning_rate": 3.6506058597274556e-06,
+      "loss": 0.4019,
+      "step": 9409
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5736335716129107,
+      "learning_rate": 3.6496125799775113e-06,
+      "loss": 0.4804,
+      "step": 9410
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7305932920688418,
+      "learning_rate": 3.648619357709243e-06,
+      "loss": 0.4882,
+      "step": 9411
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6085317276675052,
+      "learning_rate": 3.6476261929649316e-06,
+      "loss": 0.4291,
+      "step": 9412
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6984835121280041,
+      "learning_rate": 3.646633085786851e-06,
+      "loss": 0.508,
+      "step": 9413
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8424748932724646,
+      "learning_rate": 3.6456400362172752e-06,
+      "loss": 0.5327,
+      "step": 9414
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7425466890467683,
+      "learning_rate": 3.644647044298475e-06,
+      "loss": 0.5255,
+      "step": 9415
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7057164369499038,
+      "learning_rate": 3.643654110072719e-06,
+      "loss": 0.5572,
+      "step": 9416
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6238592867457389,
+      "learning_rate": 3.6426612335822735e-06,
+      "loss": 0.4244,
+      "step": 9417
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6145404264522862,
+      "learning_rate": 3.641668414869399e-06,
+      "loss": 0.4653,
+      "step": 9418
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7842686843129191,
+      "learning_rate": 3.6406756539763633e-06,
+      "loss": 0.5227,
+      "step": 9419
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8000686490823119,
+      "learning_rate": 3.6396829509454203e-06,
+      "loss": 0.4634,
+      "step": 9420
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6307034047154679,
+      "learning_rate": 3.638690305818826e-06,
+      "loss": 0.4187,
+      "step": 9421
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6217178830330693,
+      "learning_rate": 3.637697718638837e-06,
+      "loss": 0.4792,
+      "step": 9422
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7085604471968104,
+      "learning_rate": 3.6367051894477017e-06,
+      "loss": 0.4821,
+      "step": 9423
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.624567134615195,
+      "learning_rate": 3.6357127182876705e-06,
+      "loss": 0.4685,
+      "step": 9424
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.583034659519317,
+      "learning_rate": 3.6347203052009895e-06,
+      "loss": 0.4229,
+      "step": 9425
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5609245545654077,
+      "learning_rate": 3.6337279502299017e-06,
+      "loss": 0.4346,
+      "step": 9426
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6057397309858307,
+      "learning_rate": 3.632735653416649e-06,
+      "loss": 0.4746,
+      "step": 9427
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7008415882385254,
+      "learning_rate": 3.6317434148034714e-06,
+      "loss": 0.4976,
+      "step": 9428
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.600971276732898,
+      "learning_rate": 3.6307512344326034e-06,
+      "loss": 0.4568,
+      "step": 9429
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6374964222218062,
+      "learning_rate": 3.629759112346278e-06,
+      "loss": 0.3921,
+      "step": 9430
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6207756351778158,
+      "learning_rate": 3.6287670485867306e-06,
+      "loss": 0.4926,
+      "step": 9431
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.854422474203202,
+      "learning_rate": 3.6277750431961877e-06,
+      "loss": 0.5187,
+      "step": 9432
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9067806553141458,
+      "learning_rate": 3.626783096216877e-06,
+      "loss": 0.5027,
+      "step": 9433
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7618917669033753,
+      "learning_rate": 3.6257912076910214e-06,
+      "loss": 0.5116,
+      "step": 9434
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7451529824388118,
+      "learning_rate": 3.6247993776608426e-06,
+      "loss": 0.4833,
+      "step": 9435
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5961462216264772,
+      "learning_rate": 3.6238076061685593e-06,
+      "loss": 0.4642,
+      "step": 9436
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7265659719787936,
+      "learning_rate": 3.6228158932563895e-06,
+      "loss": 0.5014,
+      "step": 9437
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7626185257371304,
+      "learning_rate": 3.6218242389665457e-06,
+      "loss": 0.5255,
+      "step": 9438
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8052022781133142,
+      "learning_rate": 3.62083264334124e-06,
+      "loss": 0.5387,
+      "step": 9439
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7209810536499134,
+      "learning_rate": 3.619841106422682e-06,
+      "loss": 0.5311,
+      "step": 9440
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7741278287168436,
+      "learning_rate": 3.6188496282530774e-06,
+      "loss": 0.481,
+      "step": 9441
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7068216646728024,
+      "learning_rate": 3.6178582088746307e-06,
+      "loss": 0.4893,
+      "step": 9442
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8401408212994929,
+      "learning_rate": 3.616866848329542e-06,
+      "loss": 0.4868,
+      "step": 9443
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7424489850453143,
+      "learning_rate": 3.615875546660013e-06,
+      "loss": 0.4622,
+      "step": 9444
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5707105923105024,
+      "learning_rate": 3.6148843039082394e-06,
+      "loss": 0.4654,
+      "step": 9445
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5480934382977639,
+      "learning_rate": 3.613893120116413e-06,
+      "loss": 0.4318,
+      "step": 9446
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6719126192826061,
+      "learning_rate": 3.6129019953267285e-06,
+      "loss": 0.4529,
+      "step": 9447
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.577864508267278,
+      "learning_rate": 3.611910929581373e-06,
+      "loss": 0.4608,
+      "step": 9448
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7205605247021227,
+      "learning_rate": 3.6109199229225346e-06,
+      "loss": 0.4669,
+      "step": 9449
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9227561583055726,
+      "learning_rate": 3.609928975392395e-06,
+      "loss": 0.5241,
+      "step": 9450
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8264774168956006,
+      "learning_rate": 3.6089380870331377e-06,
+      "loss": 0.5459,
+      "step": 9451
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6348407077098837,
+      "learning_rate": 3.6079472578869405e-06,
+      "loss": 0.4165,
+      "step": 9452
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6035479530103958,
+      "learning_rate": 3.6069564879959805e-06,
+      "loss": 0.4698,
+      "step": 9453
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0151899137145066,
+      "learning_rate": 3.605965777402431e-06,
+      "loss": 0.5063,
+      "step": 9454
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7272700147929374,
+      "learning_rate": 3.6049751261484624e-06,
+      "loss": 0.5454,
+      "step": 9455
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8989338736016839,
+      "learning_rate": 3.6039845342762464e-06,
+      "loss": 0.5116,
+      "step": 9456
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6306512603646491,
+      "learning_rate": 3.6029940018279474e-06,
+      "loss": 0.4634,
+      "step": 9457
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6193870711223556,
+      "learning_rate": 3.60200352884573e-06,
+      "loss": 0.5318,
+      "step": 9458
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5828212196482788,
+      "learning_rate": 3.6010131153717544e-06,
+      "loss": 0.4574,
+      "step": 9459
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5727501490341228,
+      "learning_rate": 3.6000227614481797e-06,
+      "loss": 0.5265,
+      "step": 9460
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.678357157869081,
+      "learning_rate": 3.599032467117163e-06,
+      "loss": 0.5516,
+      "step": 9461
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5644482035468292,
+      "learning_rate": 3.5980422324208576e-06,
+      "loss": 0.4234,
+      "step": 9462
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6079344633391223,
+      "learning_rate": 3.597052057401414e-06,
+      "loss": 0.4265,
+      "step": 9463
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5825564950202775,
+      "learning_rate": 3.596061942100981e-06,
+      "loss": 0.4027,
+      "step": 9464
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6088844888324194,
+      "learning_rate": 3.5950718865617052e-06,
+      "loss": 0.4967,
+      "step": 9465
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6896780299535039,
+      "learning_rate": 3.594081890825729e-06,
+      "loss": 0.4752,
+      "step": 9466
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7094657172600143,
+      "learning_rate": 3.593091954935194e-06,
+      "loss": 0.5189,
+      "step": 9467
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7255489766048862,
+      "learning_rate": 3.592102078932237e-06,
+      "loss": 0.6105,
+      "step": 9468
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5335851202317338,
+      "learning_rate": 3.5911122628589977e-06,
+      "loss": 0.4162,
+      "step": 9469
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.958267793174056,
+      "learning_rate": 3.5901225067576073e-06,
+      "loss": 0.469,
+      "step": 9470
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8552163739072568,
+      "learning_rate": 3.589132810670196e-06,
+      "loss": 0.5468,
+      "step": 9471
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.1334336176113844,
+      "learning_rate": 3.588143174638893e-06,
+      "loss": 0.5105,
+      "step": 9472
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6796844741501856,
+      "learning_rate": 3.5871535987058226e-06,
+      "loss": 0.4622,
+      "step": 9473
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6019998325886428,
+      "learning_rate": 3.5861640829131096e-06,
+      "loss": 0.4492,
+      "step": 9474
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6671691601084976,
+      "learning_rate": 3.585174627302873e-06,
+      "loss": 0.4612,
+      "step": 9475
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6137077765714484,
+      "learning_rate": 3.584185231917231e-06,
+      "loss": 0.4189,
+      "step": 9476
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6408745021382739,
+      "learning_rate": 3.5831958967983004e-06,
+      "loss": 0.4773,
+      "step": 9477
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.9151327707390334,
+      "learning_rate": 3.5822066219881922e-06,
+      "loss": 0.5135,
+      "step": 9478
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7123509014075077,
+      "learning_rate": 3.5812174075290185e-06,
+      "loss": 0.4742,
+      "step": 9479
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6151227884910276,
+      "learning_rate": 3.580228253462883e-06,
+      "loss": 0.4843,
+      "step": 9480
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6466972396294765,
+      "learning_rate": 3.579239159831896e-06,
+      "loss": 0.494,
+      "step": 9481
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5525207693083822,
+      "learning_rate": 3.5782501266781577e-06,
+      "loss": 0.4292,
+      "step": 9482
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7272640659632693,
+      "learning_rate": 3.5772611540437687e-06,
+      "loss": 0.476,
+      "step": 9483
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6582388715087351,
+      "learning_rate": 3.5762722419708255e-06,
+      "loss": 0.4672,
+      "step": 9484
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.0332531210569487,
+      "learning_rate": 3.575283390501424e-06,
+      "loss": 0.5439,
+      "step": 9485
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7006529815804571,
+      "learning_rate": 3.574294599677656e-06,
+      "loss": 0.5591,
+      "step": 9486
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.9239951531892011,
+      "learning_rate": 3.57330586954161e-06,
+      "loss": 0.5313,
+      "step": 9487
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6797135075251783,
+      "learning_rate": 3.5723172001353747e-06,
+      "loss": 0.4539,
+      "step": 9488
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.9071326519953101,
+      "learning_rate": 3.571328591501033e-06,
+      "loss": 0.5236,
+      "step": 9489
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8655001204343843,
+      "learning_rate": 3.570340043680669e-06,
+      "loss": 0.4913,
+      "step": 9490
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.547057778314284,
+      "learning_rate": 3.5693515567163604e-06,
+      "loss": 0.4886,
+      "step": 9491
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7234796173785958,
+      "learning_rate": 3.568363130650182e-06,
+      "loss": 0.4973,
+      "step": 9492
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5463726485089478,
+      "learning_rate": 3.567374765524213e-06,
+      "loss": 0.4064,
+      "step": 9493
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.708796163515261,
+      "learning_rate": 3.5663864613805217e-06,
+      "loss": 0.5194,
+      "step": 9494
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8742522625023077,
+      "learning_rate": 3.565398218261178e-06,
+      "loss": 0.5662,
+      "step": 9495
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7488353337072939,
+      "learning_rate": 3.564410036208247e-06,
+      "loss": 0.4824,
+      "step": 9496
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6227279932724672,
+      "learning_rate": 3.563421915263794e-06,
+      "loss": 0.407,
+      "step": 9497
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7032691500929044,
+      "learning_rate": 3.562433855469879e-06,
+      "loss": 0.4756,
+      "step": 9498
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7684014194535747,
+      "learning_rate": 3.561445856868561e-06,
+      "loss": 0.5294,
+      "step": 9499
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6712242771991961,
+      "learning_rate": 3.5604579195018962e-06,
+      "loss": 0.4962,
+      "step": 9500
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7469355859788023,
+      "learning_rate": 3.5594700434119378e-06,
+      "loss": 0.4495,
+      "step": 9501
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6225801155766425,
+      "learning_rate": 3.558482228640736e-06,
+      "loss": 0.4564,
+      "step": 9502
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.692915274564329,
+      "learning_rate": 3.5574944752303394e-06,
+      "loss": 0.4773,
+      "step": 9503
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7203387464232403,
+      "learning_rate": 3.5565067832227933e-06,
+      "loss": 0.5009,
+      "step": 9504
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6163578035778042,
+      "learning_rate": 3.5555191526601395e-06,
+      "loss": 0.5272,
+      "step": 9505
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7545949123118146,
+      "learning_rate": 3.554531583584421e-06,
+      "loss": 0.5605,
+      "step": 9506
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7096062047494669,
+      "learning_rate": 3.5535440760376736e-06,
+      "loss": 0.5372,
+      "step": 9507
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7589371207711612,
+      "learning_rate": 3.5525566300619318e-06,
+      "loss": 0.5539,
+      "step": 9508
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.740793068235596,
+      "learning_rate": 3.55156924569923e-06,
+      "loss": 0.4934,
+      "step": 9509
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7266898947552103,
+      "learning_rate": 3.5505819229915963e-06,
+      "loss": 0.5067,
+      "step": 9510
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.676615527553606,
+      "learning_rate": 3.549594661981059e-06,
+      "loss": 0.4779,
+      "step": 9511
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7122453572063476,
+      "learning_rate": 3.5486074627096425e-06,
+      "loss": 0.4317,
+      "step": 9512
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6194319823175132,
+      "learning_rate": 3.5476203252193676e-06,
+      "loss": 0.5079,
+      "step": 9513
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.59752804298602,
+      "learning_rate": 3.5466332495522543e-06,
+      "loss": 0.4442,
+      "step": 9514
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6544476110689336,
+      "learning_rate": 3.5456462357503197e-06,
+      "loss": 0.4124,
+      "step": 9515
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6783805184502651,
+      "learning_rate": 3.5446592838555773e-06,
+      "loss": 0.4609,
+      "step": 9516
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5867172345646141,
+      "learning_rate": 3.543672393910037e-06,
+      "loss": 0.4424,
+      "step": 9517
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6976572604775314,
+      "learning_rate": 3.5426855659557107e-06,
+      "loss": 0.5671,
+      "step": 9518
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7744659246456838,
+      "learning_rate": 3.541698800034603e-06,
+      "loss": 0.5592,
+      "step": 9519
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7010569671479242,
+      "learning_rate": 3.540712096188717e-06,
+      "loss": 0.5154,
+      "step": 9520
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6287833384926903,
+      "learning_rate": 3.5397254544600545e-06,
+      "loss": 0.4634,
+      "step": 9521
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6728368009414035,
+      "learning_rate": 3.5387388748906122e-06,
+      "loss": 0.5306,
+      "step": 9522
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.9248970475385573,
+      "learning_rate": 3.537752357522387e-06,
+      "loss": 0.5508,
+      "step": 9523
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8055062852865371,
+      "learning_rate": 3.536765902397371e-06,
+      "loss": 0.557,
+      "step": 9524
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6662121900971837,
+      "learning_rate": 3.535779509557555e-06,
+      "loss": 0.4929,
+      "step": 9525
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6850244133045399,
+      "learning_rate": 3.5347931790449257e-06,
+      "loss": 0.4481,
+      "step": 9526
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7584437978991218,
+      "learning_rate": 3.5338069109014694e-06,
+      "loss": 0.5271,
+      "step": 9527
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8100403308718558,
+      "learning_rate": 3.5328207051691672e-06,
+      "loss": 0.5173,
+      "step": 9528
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8487650941987739,
+      "learning_rate": 3.5318345618899992e-06,
+      "loss": 0.5182,
+      "step": 9529
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6133629067743392,
+      "learning_rate": 3.5308484811059402e-06,
+      "loss": 0.4626,
+      "step": 9530
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.942036253091359,
+      "learning_rate": 3.5298624628589683e-06,
+      "loss": 0.5166,
+      "step": 9531
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6445150671411881,
+      "learning_rate": 3.528876507191055e-06,
+      "loss": 0.4719,
+      "step": 9532
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5909108258872108,
+      "learning_rate": 3.527890614144166e-06,
+      "loss": 0.4,
+      "step": 9533
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5798541954088046,
+      "learning_rate": 3.52690478376027e-06,
+      "loss": 0.4077,
+      "step": 9534
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6778919284185424,
+      "learning_rate": 3.52591901608133e-06,
+      "loss": 0.502,
+      "step": 9535
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.820877798191278,
+      "learning_rate": 3.5249333111493066e-06,
+      "loss": 0.5297,
+      "step": 9536
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.697150675883807,
+      "learning_rate": 3.523947669006159e-06,
+      "loss": 0.4982,
+      "step": 9537
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6344091185942939,
+      "learning_rate": 3.522962089693843e-06,
+      "loss": 0.4921,
+      "step": 9538
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.3954342319233473,
+      "learning_rate": 3.5219765732543097e-06,
+      "loss": 0.5411,
+      "step": 9539
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7648751669554349,
+      "learning_rate": 3.520991119729511e-06,
+      "loss": 0.5726,
+      "step": 9540
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7401215041280158,
+      "learning_rate": 3.520005729161394e-06,
+      "loss": 0.5643,
+      "step": 9541
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6102587041592459,
+      "learning_rate": 3.5190204015919023e-06,
+      "loss": 0.4681,
+      "step": 9542
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6515046034696517,
+      "learning_rate": 3.518035137062981e-06,
+      "loss": 0.4295,
+      "step": 9543
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6335726288547892,
+      "learning_rate": 3.5170499356165688e-06,
+      "loss": 0.4862,
+      "step": 9544
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.931279516981549,
+      "learning_rate": 3.5160647972946016e-06,
+      "loss": 0.4985,
+      "step": 9545
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6541495643489148,
+      "learning_rate": 3.5150797221390137e-06,
+      "loss": 0.4314,
+      "step": 9546
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6574708678092356,
+      "learning_rate": 3.514094710191738e-06,
+      "loss": 0.4937,
+      "step": 9547
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7812278288880611,
+      "learning_rate": 3.5131097614947007e-06,
+      "loss": 0.4697,
+      "step": 9548
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6514992822042747,
+      "learning_rate": 3.51212487608983e-06,
+      "loss": 0.4267,
+      "step": 9549
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6239247188201337,
+      "learning_rate": 3.5111400540190493e-06,
+      "loss": 0.4916,
+      "step": 9550
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.87093789534363,
+      "learning_rate": 3.5101552953242785e-06,
+      "loss": 0.5558,
+      "step": 9551
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7011620251504134,
+      "learning_rate": 3.509170600047436e-06,
+      "loss": 0.5146,
+      "step": 9552
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7252922378649093,
+      "learning_rate": 3.5081859682304366e-06,
+      "loss": 0.5136,
+      "step": 9553
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6490925619006583,
+      "learning_rate": 3.5072013999151917e-06,
+      "loss": 0.4826,
+      "step": 9554
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6454691033825337,
+      "learning_rate": 3.506216895143615e-06,
+      "loss": 0.4714,
+      "step": 9555
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6220440688297076,
+      "learning_rate": 3.505232453957611e-06,
+      "loss": 0.4995,
+      "step": 9556
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.793677132118531,
+      "learning_rate": 3.5042480763990857e-06,
+      "loss": 0.5655,
+      "step": 9557
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8110021035452885,
+      "learning_rate": 3.5032637625099397e-06,
+      "loss": 0.5614,
+      "step": 9558
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8370877753247526,
+      "learning_rate": 3.5022795123320718e-06,
+      "loss": 0.5392,
+      "step": 9559
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7992332193530127,
+      "learning_rate": 3.50129532590738e-06,
+      "loss": 0.5056,
+      "step": 9560
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6268060304286129,
+      "learning_rate": 3.5003112032777563e-06,
+      "loss": 0.4374,
+      "step": 9561
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7427240184302516,
+      "learning_rate": 3.499327144485093e-06,
+      "loss": 0.5067,
+      "step": 9562
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6031566811450092,
+      "learning_rate": 3.4983431495712773e-06,
+      "loss": 0.4219,
+      "step": 9563
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.692715235767375,
+      "learning_rate": 3.4973592185781955e-06,
+      "loss": 0.4432,
+      "step": 9564
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6455174718618477,
+      "learning_rate": 3.4963753515477294e-06,
+      "loss": 0.4991,
+      "step": 9565
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6618973069986885,
+      "learning_rate": 3.4953915485217603e-06,
+      "loss": 0.4734,
+      "step": 9566
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6091413365808238,
+      "learning_rate": 3.4944078095421636e-06,
+      "loss": 0.4451,
+      "step": 9567
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6361375464784068,
+      "learning_rate": 3.4934241346508165e-06,
+      "loss": 0.4555,
+      "step": 9568
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5552947658644004,
+      "learning_rate": 3.492440523889589e-06,
+      "loss": 0.437,
+      "step": 9569
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.638820309850292,
+      "learning_rate": 3.4914569773003503e-06,
+      "loss": 0.4663,
+      "step": 9570
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6409077448931367,
+      "learning_rate": 3.4904734949249685e-06,
+      "loss": 0.4226,
+      "step": 9571
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6524659815371653,
+      "learning_rate": 3.489490076805306e-06,
+      "loss": 0.4889,
+      "step": 9572
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8104513560359653,
+      "learning_rate": 3.4885067229832246e-06,
+      "loss": 0.5174,
+      "step": 9573
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.573260450430759,
+      "learning_rate": 3.487523433500582e-06,
+      "loss": 0.4272,
+      "step": 9574
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5968621977917752,
+      "learning_rate": 3.4865402083992335e-06,
+      "loss": 0.4538,
+      "step": 9575
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.753852619431407,
+      "learning_rate": 3.485557047721032e-06,
+      "loss": 0.3875,
+      "step": 9576
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6648529597676575,
+      "learning_rate": 3.484573951507827e-06,
+      "loss": 0.4558,
+      "step": 9577
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6038583918411186,
+      "learning_rate": 3.4835909198014674e-06,
+      "loss": 0.4875,
+      "step": 9578
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6562597198557258,
+      "learning_rate": 3.482607952643794e-06,
+      "loss": 0.4694,
+      "step": 9579
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.637734725565027,
+      "learning_rate": 3.4816250500766534e-06,
+      "loss": 0.4651,
+      "step": 9580
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6243187144183077,
+      "learning_rate": 3.480642212141882e-06,
+      "loss": 0.4524,
+      "step": 9581
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6205104751826146,
+      "learning_rate": 3.4796594388813165e-06,
+      "loss": 0.4633,
+      "step": 9582
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7537233492004364,
+      "learning_rate": 3.4786767303367904e-06,
+      "loss": 0.5545,
+      "step": 9583
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8048135126357497,
+      "learning_rate": 3.477694086550134e-06,
+      "loss": 0.5571,
+      "step": 9584
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7253749724185615,
+      "learning_rate": 3.476711507563176e-06,
+      "loss": 0.5059,
+      "step": 9585
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.836737129944647,
+      "learning_rate": 3.475728993417742e-06,
+      "loss": 0.5169,
+      "step": 9586
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7551192186028224,
+      "learning_rate": 3.474746544155653e-06,
+      "loss": 0.4917,
+      "step": 9587
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7168967448397087,
+      "learning_rate": 3.47376415981873e-06,
+      "loss": 0.445,
+      "step": 9588
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5533768624867513,
+      "learning_rate": 3.47278184044879e-06,
+      "loss": 0.4699,
+      "step": 9589
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6531325892846266,
+      "learning_rate": 3.4717995860876462e-06,
+      "loss": 0.4578,
+      "step": 9590
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7436235037554774,
+      "learning_rate": 3.4708173967771107e-06,
+      "loss": 0.488,
+      "step": 9591
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8112057817512172,
+      "learning_rate": 3.46983527255899e-06,
+      "loss": 0.5133,
+      "step": 9592
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6072924295617089,
+      "learning_rate": 3.468853213475094e-06,
+      "loss": 0.4526,
+      "step": 9593
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6088393253612352,
+      "learning_rate": 3.467871219567224e-06,
+      "loss": 0.4677,
+      "step": 9594
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6354448602537883,
+      "learning_rate": 3.4668892908771802e-06,
+      "loss": 0.4607,
+      "step": 9595
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.588668623792249,
+      "learning_rate": 3.4659074274467596e-06,
+      "loss": 0.4394,
+      "step": 9596
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6483942601325419,
+      "learning_rate": 3.464925629317758e-06,
+      "loss": 0.4328,
+      "step": 9597
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6266912586228107,
+      "learning_rate": 3.463943896531967e-06,
+      "loss": 0.4791,
+      "step": 9598
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6582908632337373,
+      "learning_rate": 3.462962229131176e-06,
+      "loss": 0.4797,
+      "step": 9599
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8104118433322373,
+      "learning_rate": 3.461980627157171e-06,
+      "loss": 0.4799,
+      "step": 9600
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6214936350332427,
+      "learning_rate": 3.4609990906517367e-06,
+      "loss": 0.47,
+      "step": 9601
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8602054275217241,
+      "learning_rate": 3.460017619656652e-06,
+      "loss": 0.5482,
+      "step": 9602
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8458931340294048,
+      "learning_rate": 3.459036214213697e-06,
+      "loss": 0.5587,
+      "step": 9603
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7366719521818139,
+      "learning_rate": 3.458054874364643e-06,
+      "loss": 0.531,
+      "step": 9604
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7087381750490165,
+      "learning_rate": 3.4570736001512685e-06,
+      "loss": 0.5459,
+      "step": 9605
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7783415464754193,
+      "learning_rate": 3.45609239161534e-06,
+      "loss": 0.4991,
+      "step": 9606
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6350011996685322,
+      "learning_rate": 3.455111248798625e-06,
+      "loss": 0.481,
+      "step": 9607
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7195693241568325,
+      "learning_rate": 3.4541301717428877e-06,
+      "loss": 0.4741,
+      "step": 9608
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7088978698882897,
+      "learning_rate": 3.4531491604898888e-06,
+      "loss": 0.464,
+      "step": 9609
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.721564441042376,
+      "learning_rate": 3.452168215081387e-06,
+      "loss": 0.5201,
+      "step": 9610
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7811436353412271,
+      "learning_rate": 3.451187335559138e-06,
+      "loss": 0.5216,
+      "step": 9611
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6132511808358145,
+      "learning_rate": 3.450206521964895e-06,
+      "loss": 0.4325,
+      "step": 9612
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6925964234648885,
+      "learning_rate": 3.4492257743404078e-06,
+      "loss": 0.4474,
+      "step": 9613
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6562312598877206,
+      "learning_rate": 3.4482450927274242e-06,
+      "loss": 0.4717,
+      "step": 9614
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.0228066324862592,
+      "learning_rate": 3.4472644771676878e-06,
+      "loss": 0.5333,
+      "step": 9615
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5807906563440279,
+      "learning_rate": 3.4462839277029393e-06,
+      "loss": 0.4398,
+      "step": 9616
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6340389532708894,
+      "learning_rate": 3.445303444374921e-06,
+      "loss": 0.4507,
+      "step": 9617
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7427823046939891,
+      "learning_rate": 3.4443230272253662e-06,
+      "loss": 0.5257,
+      "step": 9618
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6269877363995071,
+      "learning_rate": 3.44334267629601e-06,
+      "loss": 0.477,
+      "step": 9619
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.743625191470743,
+      "learning_rate": 3.442362391628581e-06,
+      "loss": 0.5226,
+      "step": 9620
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6038063353448906,
+      "learning_rate": 3.4413821732648077e-06,
+      "loss": 0.4348,
+      "step": 9621
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7309311624728724,
+      "learning_rate": 3.4404020212464147e-06,
+      "loss": 0.4612,
+      "step": 9622
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6570554203276916,
+      "learning_rate": 3.439421935615124e-06,
+      "loss": 0.4936,
+      "step": 9623
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7456937621873247,
+      "learning_rate": 3.4384419164126537e-06,
+      "loss": 0.5251,
+      "step": 9624
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6025230415683878,
+      "learning_rate": 3.4374619636807217e-06,
+      "loss": 0.4733,
+      "step": 9625
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7239543169751721,
+      "learning_rate": 3.4364820774610406e-06,
+      "loss": 0.5405,
+      "step": 9626
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6155250039409523,
+      "learning_rate": 3.4355022577953216e-06,
+      "loss": 0.4676,
+      "step": 9627
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.212115193194442,
+      "learning_rate": 3.4345225047252718e-06,
+      "loss": 0.515,
+      "step": 9628
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5883823266749431,
+      "learning_rate": 3.4335428182925956e-06,
+      "loss": 0.4876,
+      "step": 9629
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6486140846555637,
+      "learning_rate": 3.4325631985389964e-06,
+      "loss": 0.4339,
+      "step": 9630
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6670372602311031,
+      "learning_rate": 3.4315836455061735e-06,
+      "loss": 0.4921,
+      "step": 9631
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6278492413799993,
+      "learning_rate": 3.4306041592358218e-06,
+      "loss": 0.5041,
+      "step": 9632
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5915076063710176,
+      "learning_rate": 3.429624739769637e-06,
+      "loss": 0.465,
+      "step": 9633
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7136725725353825,
+      "learning_rate": 3.4286453871493087e-06,
+      "loss": 0.5453,
+      "step": 9634
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.3624164864499388,
+      "learning_rate": 3.4276661014165246e-06,
+      "loss": 0.5368,
+      "step": 9635
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.747894990043783,
+      "learning_rate": 3.426686882612971e-06,
+      "loss": 0.5623,
+      "step": 9636
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7980325720582407,
+      "learning_rate": 3.4257077307803286e-06,
+      "loss": 0.562,
+      "step": 9637
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.712089915023937,
+      "learning_rate": 3.4247286459602775e-06,
+      "loss": 0.4962,
+      "step": 9638
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5620463504382522,
+      "learning_rate": 3.4237496281944936e-06,
+      "loss": 0.3795,
+      "step": 9639
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6172752564627064,
+      "learning_rate": 3.4227706775246517e-06,
+      "loss": 0.4245,
+      "step": 9640
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.609990838409253,
+      "learning_rate": 3.4217917939924204e-06,
+      "loss": 0.5065,
+      "step": 9641
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8053919550094233,
+      "learning_rate": 3.4208129776394712e-06,
+      "loss": 0.5541,
+      "step": 9642
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.120625348506504,
+      "learning_rate": 3.4198342285074667e-06,
+      "loss": 0.5251,
+      "step": 9643
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9142736406870288,
+      "learning_rate": 3.418855546638069e-06,
+      "loss": 0.4557,
+      "step": 9644
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5845516104725537,
+      "learning_rate": 3.417876932072939e-06,
+      "loss": 0.4323,
+      "step": 9645
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7049217338081998,
+      "learning_rate": 3.4168983848537316e-06,
+      "loss": 0.4971,
+      "step": 9646
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7026243089837511,
+      "learning_rate": 3.4159199050221015e-06,
+      "loss": 0.4888,
+      "step": 9647
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6423647309746382,
+      "learning_rate": 3.414941492619699e-06,
+      "loss": 0.4698,
+      "step": 9648
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6964945513856139,
+      "learning_rate": 3.413963147688172e-06,
+      "loss": 0.5212,
+      "step": 9649
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6214252614056979,
+      "learning_rate": 3.4129848702691648e-06,
+      "loss": 0.4758,
+      "step": 9650
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7473654028416774,
+      "learning_rate": 3.412006660404321e-06,
+      "loss": 0.4981,
+      "step": 9651
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7338716613941693,
+      "learning_rate": 3.4110285181352787e-06,
+      "loss": 0.4483,
+      "step": 9652
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5979132549143013,
+      "learning_rate": 3.410050443503675e-06,
+      "loss": 0.4381,
+      "step": 9653
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6845570601229272,
+      "learning_rate": 3.409072436551142e-06,
+      "loss": 0.4958,
+      "step": 9654
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5776470081109305,
+      "learning_rate": 3.4080944973193127e-06,
+      "loss": 0.4978,
+      "step": 9655
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6818693753463683,
+      "learning_rate": 3.4071166258498134e-06,
+      "loss": 0.5582,
+      "step": 9656
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7405322976684284,
+      "learning_rate": 3.4061388221842694e-06,
+      "loss": 0.5394,
+      "step": 9657
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7630809837510244,
+      "learning_rate": 3.4051610863643026e-06,
+      "loss": 0.5184,
+      "step": 9658
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6240005113126648,
+      "learning_rate": 3.4041834184315325e-06,
+      "loss": 0.5185,
+      "step": 9659
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.738748319660092,
+      "learning_rate": 3.403205818427574e-06,
+      "loss": 0.5198,
+      "step": 9660
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.85820358644243,
+      "learning_rate": 3.4022282863940415e-06,
+      "loss": 0.5498,
+      "step": 9661
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7723464552274426,
+      "learning_rate": 3.401250822372545e-06,
+      "loss": 0.5401,
+      "step": 9662
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9412500161288903,
+      "learning_rate": 3.4002734264046923e-06,
+      "loss": 0.4923,
+      "step": 9663
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6688416186565718,
+      "learning_rate": 3.3992960985320877e-06,
+      "loss": 0.4637,
+      "step": 9664
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6901652374288079,
+      "learning_rate": 3.3983188387963334e-06,
+      "loss": 0.4753,
+      "step": 9665
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6530480862320649,
+      "learning_rate": 3.397341647239025e-06,
+      "loss": 0.5008,
+      "step": 9666
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7884568324014779,
+      "learning_rate": 3.3963645239017644e-06,
+      "loss": 0.5431,
+      "step": 9667
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8573859812149741,
+      "learning_rate": 3.3953874688261413e-06,
+      "loss": 0.49,
+      "step": 9668
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.2110211371568325,
+      "learning_rate": 3.3944104820537453e-06,
+      "loss": 0.5374,
+      "step": 9669
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6887016607639445,
+      "learning_rate": 3.393433563626165e-06,
+      "loss": 0.4968,
+      "step": 9670
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.754896727447843,
+      "learning_rate": 3.392456713584984e-06,
+      "loss": 0.5469,
+      "step": 9671
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7532981159357144,
+      "learning_rate": 3.3914799319717837e-06,
+      "loss": 0.5089,
+      "step": 9672
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7393912423406087,
+      "learning_rate": 3.390503218828143e-06,
+      "loss": 0.5416,
+      "step": 9673
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8085606629083892,
+      "learning_rate": 3.389526574195636e-06,
+      "loss": 0.5285,
+      "step": 9674
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6091851025976537,
+      "learning_rate": 3.388549998115837e-06,
+      "loss": 0.4232,
+      "step": 9675
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5707818476742368,
+      "learning_rate": 3.3875734906303146e-06,
+      "loss": 0.4511,
+      "step": 9676
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7227538980858315,
+      "learning_rate": 3.386597051780637e-06,
+      "loss": 0.4839,
+      "step": 9677
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6104563736198384,
+      "learning_rate": 3.3856206816083647e-06,
+      "loss": 0.4512,
+      "step": 9678
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7664101677574381,
+      "learning_rate": 3.384644380155063e-06,
+      "loss": 0.4882,
+      "step": 9679
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7209889836951127,
+      "learning_rate": 3.383668147462289e-06,
+      "loss": 0.5079,
+      "step": 9680
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.668217285319877,
+      "learning_rate": 3.3826919835715965e-06,
+      "loss": 0.4372,
+      "step": 9681
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6373655781152588,
+      "learning_rate": 3.3817158885245376e-06,
+      "loss": 0.5166,
+      "step": 9682
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9074841743386916,
+      "learning_rate": 3.380739862362662e-06,
+      "loss": 0.5054,
+      "step": 9683
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7556160265804783,
+      "learning_rate": 3.3797639051275166e-06,
+      "loss": 0.5402,
+      "step": 9684
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7500841576864383,
+      "learning_rate": 3.3787880168606442e-06,
+      "loss": 0.4938,
+      "step": 9685
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6651220283663039,
+      "learning_rate": 3.377812197603585e-06,
+      "loss": 0.4682,
+      "step": 9686
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6644771417163498,
+      "learning_rate": 3.3768364473978766e-06,
+      "loss": 0.4449,
+      "step": 9687
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5885818079786101,
+      "learning_rate": 3.375860766285054e-06,
+      "loss": 0.438,
+      "step": 9688
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6493892725602527,
+      "learning_rate": 3.3748851543066486e-06,
+      "loss": 0.4498,
+      "step": 9689
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5903591044995382,
+      "learning_rate": 3.373909611504189e-06,
+      "loss": 0.4344,
+      "step": 9690
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8475285021136163,
+      "learning_rate": 3.3729341379192006e-06,
+      "loss": 0.4975,
+      "step": 9691
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6043653900253609,
+      "learning_rate": 3.371958733593207e-06,
+      "loss": 0.4319,
+      "step": 9692
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5772567621284745,
+      "learning_rate": 3.370983398567727e-06,
+      "loss": 0.4342,
+      "step": 9693
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6100995651157373,
+      "learning_rate": 3.370008132884278e-06,
+      "loss": 0.4757,
+      "step": 9694
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5980289280900019,
+      "learning_rate": 3.3690329365843743e-06,
+      "loss": 0.4743,
+      "step": 9695
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8002261870219475,
+      "learning_rate": 3.3680578097095275e-06,
+      "loss": 0.4867,
+      "step": 9696
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7843917144754352,
+      "learning_rate": 3.3670827523012443e-06,
+      "loss": 0.4666,
+      "step": 9697
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8952009648199384,
+      "learning_rate": 3.3661077644010308e-06,
+      "loss": 0.5205,
+      "step": 9698
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.006639819442437,
+      "learning_rate": 3.3651328460503884e-06,
+      "loss": 0.5611,
+      "step": 9699
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8166682729650561,
+      "learning_rate": 3.364157997290816e-06,
+      "loss": 0.5293,
+      "step": 9700
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6763966588639446,
+      "learning_rate": 3.363183218163811e-06,
+      "loss": 0.5146,
+      "step": 9701
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7478200352706385,
+      "learning_rate": 3.3622085087108653e-06,
+      "loss": 0.5068,
+      "step": 9702
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.674491227335443,
+      "learning_rate": 3.3612338689734693e-06,
+      "loss": 0.444,
+      "step": 9703
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.618147605645662,
+      "learning_rate": 3.360259298993111e-06,
+      "loss": 0.4791,
+      "step": 9704
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8377174390100891,
+      "learning_rate": 3.359284798811275e-06,
+      "loss": 0.5463,
+      "step": 9705
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5628948708384927,
+      "learning_rate": 3.3583103684694418e-06,
+      "loss": 0.4061,
+      "step": 9706
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.654613670938272,
+      "learning_rate": 3.35733600800909e-06,
+      "loss": 0.3981,
+      "step": 9707
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6085584585489507,
+      "learning_rate": 3.3563617174716954e-06,
+      "loss": 0.4479,
+      "step": 9708
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6910659131927186,
+      "learning_rate": 3.3553874968987304e-06,
+      "loss": 0.4891,
+      "step": 9709
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7340759927453796,
+      "learning_rate": 3.354413346331664e-06,
+      "loss": 0.5156,
+      "step": 9710
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7055322321284114,
+      "learning_rate": 3.353439265811963e-06,
+      "loss": 0.5611,
+      "step": 9711
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7413780243438631,
+      "learning_rate": 3.352465255381091e-06,
+      "loss": 0.5525,
+      "step": 9712
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7415240845122817,
+      "learning_rate": 3.351491315080508e-06,
+      "loss": 0.5088,
+      "step": 9713
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6062353174867325,
+      "learning_rate": 3.350517444951672e-06,
+      "loss": 0.4666,
+      "step": 9714
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5959660815935528,
+      "learning_rate": 3.3495436450360375e-06,
+      "loss": 0.5561,
+      "step": 9715
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.795469707114058,
+      "learning_rate": 3.348569915375053e-06,
+      "loss": 0.6284,
+      "step": 9716
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.122140338181277,
+      "learning_rate": 3.3475962560101728e-06,
+      "loss": 0.5122,
+      "step": 9717
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7428464969622799,
+      "learning_rate": 3.3466226669828396e-06,
+      "loss": 0.5342,
+      "step": 9718
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5913019605528251,
+      "learning_rate": 3.345649148334496e-06,
+      "loss": 0.4627,
+      "step": 9719
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7312336653461774,
+      "learning_rate": 3.3446757001065816e-06,
+      "loss": 0.5261,
+      "step": 9720
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7278513619582827,
+      "learning_rate": 3.343702322340533e-06,
+      "loss": 0.5114,
+      "step": 9721
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7438463711809817,
+      "learning_rate": 3.3427290150777836e-06,
+      "loss": 0.548,
+      "step": 9722
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6437042828354095,
+      "learning_rate": 3.3417557783597643e-06,
+      "loss": 0.4669,
+      "step": 9723
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7360270876315178,
+      "learning_rate": 3.340782612227903e-06,
+      "loss": 0.5099,
+      "step": 9724
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.838277362309171,
+      "learning_rate": 3.339809516723623e-06,
+      "loss": 0.4796,
+      "step": 9725
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5744679583378896,
+      "learning_rate": 3.338836491888347e-06,
+      "loss": 0.4733,
+      "step": 9726
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6886928918535186,
+      "learning_rate": 3.3378635377634927e-06,
+      "loss": 0.5025,
+      "step": 9727
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8364803837127899,
+      "learning_rate": 3.336890654390475e-06,
+      "loss": 0.4694,
+      "step": 9728
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6527493447813238,
+      "learning_rate": 3.335917841810709e-06,
+      "loss": 0.435,
+      "step": 9729
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6360823743572565,
+      "learning_rate": 3.334945100065603e-06,
+      "loss": 0.4681,
+      "step": 9730
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7511257876206643,
+      "learning_rate": 3.3339724291965637e-06,
+      "loss": 0.5205,
+      "step": 9731
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6934778456312952,
+      "learning_rate": 3.332999829244994e-06,
+      "loss": 0.5217,
+      "step": 9732
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7043104721963191,
+      "learning_rate": 3.3320273002522947e-06,
+      "loss": 0.4897,
+      "step": 9733
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5824895233477628,
+      "learning_rate": 3.331054842259863e-06,
+      "loss": 0.4755,
+      "step": 9734
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5520456878069466,
+      "learning_rate": 3.3300824553090934e-06,
+      "loss": 0.4655,
+      "step": 9735
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6506681488520155,
+      "learning_rate": 3.329110139441377e-06,
+      "loss": 0.4351,
+      "step": 9736
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5785868482380239,
+      "learning_rate": 3.3281378946981034e-06,
+      "loss": 0.4369,
+      "step": 9737
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5986856199645442,
+      "learning_rate": 3.3271657211206566e-06,
+      "loss": 0.4277,
+      "step": 9738
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5732413644115761,
+      "learning_rate": 3.3261936187504195e-06,
+      "loss": 0.424,
+      "step": 9739
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6397259322602841,
+      "learning_rate": 3.3252215876287695e-06,
+      "loss": 0.4793,
+      "step": 9740
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7759409451260584,
+      "learning_rate": 3.3242496277970872e-06,
+      "loss": 0.4811,
+      "step": 9741
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6419254316919619,
+      "learning_rate": 3.3232777392967417e-06,
+      "loss": 0.4914,
+      "step": 9742
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6695568145934685,
+      "learning_rate": 3.3223059221691063e-06,
+      "loss": 0.4765,
+      "step": 9743
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6953302181373358,
+      "learning_rate": 3.3213341764555463e-06,
+      "loss": 0.5149,
+      "step": 9744
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.2427653108082664,
+      "learning_rate": 3.320362502197426e-06,
+      "loss": 0.5687,
+      "step": 9745
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6219528949720443,
+      "learning_rate": 3.319390899436107e-06,
+      "loss": 0.4848,
+      "step": 9746
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5871854366235867,
+      "learning_rate": 3.318419368212946e-06,
+      "loss": 0.4381,
+      "step": 9747
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6486799695735382,
+      "learning_rate": 3.3174479085692997e-06,
+      "loss": 0.4834,
+      "step": 9748
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7546057165321423,
+      "learning_rate": 3.3164765205465187e-06,
+      "loss": 0.5192,
+      "step": 9749
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.1427536234375142,
+      "learning_rate": 3.315505204185953e-06,
+      "loss": 0.5416,
+      "step": 9750
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.717578958570655,
+      "learning_rate": 3.314533959528948e-06,
+      "loss": 0.522,
+      "step": 9751
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7899855094305798,
+      "learning_rate": 3.3135627866168465e-06,
+      "loss": 0.5027,
+      "step": 9752
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7555347788382839,
+      "learning_rate": 3.312591685490986e-06,
+      "loss": 0.5021,
+      "step": 9753
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.591234827869523,
+      "learning_rate": 3.3116206561927074e-06,
+      "loss": 0.4164,
+      "step": 9754
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5936487336478009,
+      "learning_rate": 3.3106496987633408e-06,
+      "loss": 0.4967,
+      "step": 9755
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8057824222276049,
+      "learning_rate": 3.3096788132442186e-06,
+      "loss": 0.5461,
+      "step": 9756
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6994944841147624,
+      "learning_rate": 3.308707999676668e-06,
+      "loss": 0.5093,
+      "step": 9757
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6242269304173834,
+      "learning_rate": 3.307737258102014e-06,
+      "loss": 0.4843,
+      "step": 9758
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6617062699804608,
+      "learning_rate": 3.3067665885615762e-06,
+      "loss": 0.4927,
+      "step": 9759
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.692744029806587,
+      "learning_rate": 3.3057959910966754e-06,
+      "loss": 0.5064,
+      "step": 9760
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7759133672417525,
+      "learning_rate": 3.304825465748624e-06,
+      "loss": 0.5223,
+      "step": 9761
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7373400225765957,
+      "learning_rate": 3.303855012558736e-06,
+      "loss": 0.4861,
+      "step": 9762
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5928736253734711,
+      "learning_rate": 3.30288463156832e-06,
+      "loss": 0.4668,
+      "step": 9763
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6188388607574133,
+      "learning_rate": 3.3019143228186825e-06,
+      "loss": 0.4836,
+      "step": 9764
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5747153962949537,
+      "learning_rate": 3.300944086351124e-06,
+      "loss": 0.4496,
+      "step": 9765
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5840292329822466,
+      "learning_rate": 3.2999739222069477e-06,
+      "loss": 0.4171,
+      "step": 9766
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6775475721623886,
+      "learning_rate": 3.299003830427449e-06,
+      "loss": 0.5136,
+      "step": 9767
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8411364293719691,
+      "learning_rate": 3.2980338110539225e-06,
+      "loss": 0.495,
+      "step": 9768
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6453161014766059,
+      "learning_rate": 3.2970638641276563e-06,
+      "loss": 0.4473,
+      "step": 9769
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5595071340611592,
+      "learning_rate": 3.296093989689941e-06,
+      "loss": 0.4707,
+      "step": 9770
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7997718759840764,
+      "learning_rate": 3.2951241877820594e-06,
+      "loss": 0.5121,
+      "step": 9771
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5763401014143,
+      "learning_rate": 3.2941544584452928e-06,
+      "loss": 0.4461,
+      "step": 9772
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6561222903062404,
+      "learning_rate": 3.2931848017209202e-06,
+      "loss": 0.4714,
+      "step": 9773
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5806328696962216,
+      "learning_rate": 3.292215217650215e-06,
+      "loss": 0.47,
+      "step": 9774
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6387932215629026,
+      "learning_rate": 3.2912457062744526e-06,
+      "loss": 0.4732,
+      "step": 9775
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6183663877210778,
+      "learning_rate": 3.290276267634899e-06,
+      "loss": 0.4013,
+      "step": 9776
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6911524512861753,
+      "learning_rate": 3.289306901772822e-06,
+      "loss": 0.4614,
+      "step": 9777
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6957098179567416,
+      "learning_rate": 3.2883376087294804e-06,
+      "loss": 0.4842,
+      "step": 9778
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8892080228449,
+      "learning_rate": 3.28736838854614e-06,
+      "loss": 0.5612,
+      "step": 9779
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6049457145813931,
+      "learning_rate": 3.2863992412640546e-06,
+      "loss": 0.449,
+      "step": 9780
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6271762013024282,
+      "learning_rate": 3.285430166924477e-06,
+      "loss": 0.4191,
+      "step": 9781
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.610450083761308,
+      "learning_rate": 3.2844611655686586e-06,
+      "loss": 0.5055,
+      "step": 9782
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8269049664396605,
+      "learning_rate": 3.2834922372378465e-06,
+      "loss": 0.5623,
+      "step": 9783
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7516688504708258,
+      "learning_rate": 3.282523381973284e-06,
+      "loss": 0.5023,
+      "step": 9784
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.617465779449233,
+      "learning_rate": 3.2815545998162134e-06,
+      "loss": 0.4418,
+      "step": 9785
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7218667417046295,
+      "learning_rate": 3.280585890807872e-06,
+      "loss": 0.5132,
+      "step": 9786
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7065332850665154,
+      "learning_rate": 3.2796172549894946e-06,
+      "loss": 0.5278,
+      "step": 9787
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7764326359182376,
+      "learning_rate": 3.278648692402313e-06,
+      "loss": 0.4591,
+      "step": 9788
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8430399155602538,
+      "learning_rate": 3.2776802030875564e-06,
+      "loss": 0.5827,
+      "step": 9789
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.858285110893183,
+      "learning_rate": 3.2767117870864473e-06,
+      "loss": 0.4886,
+      "step": 9790
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6045502596390667,
+      "learning_rate": 3.275743444440213e-06,
+      "loss": 0.4728,
+      "step": 9791
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6783595771771905,
+      "learning_rate": 3.27477517519007e-06,
+      "loss": 0.5057,
+      "step": 9792
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.369519675583337,
+      "learning_rate": 3.2738069793772353e-06,
+      "loss": 0.523,
+      "step": 9793
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8034220648985635,
+      "learning_rate": 3.272838857042921e-06,
+      "loss": 0.503,
+      "step": 9794
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5758599652496723,
+      "learning_rate": 3.2718708082283367e-06,
+      "loss": 0.4371,
+      "step": 9795
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6405399937475216,
+      "learning_rate": 3.2709028329746905e-06,
+      "loss": 0.4621,
+      "step": 9796
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8655253624930225,
+      "learning_rate": 3.2699349313231854e-06,
+      "loss": 0.5093,
+      "step": 9797
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6035259467451702,
+      "learning_rate": 3.2689671033150215e-06,
+      "loss": 0.465,
+      "step": 9798
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6813695313711089,
+      "learning_rate": 3.267999348991397e-06,
+      "loss": 0.4885,
+      "step": 9799
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7322371401325936,
+      "learning_rate": 3.2670316683935047e-06,
+      "loss": 0.5128,
+      "step": 9800
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5816729474457779,
+      "learning_rate": 3.2660640615625374e-06,
+      "loss": 0.4603,
+      "step": 9801
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.621793815698565,
+      "learning_rate": 3.26509652853968e-06,
+      "loss": 0.5114,
+      "step": 9802
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6550650746031575,
+      "learning_rate": 3.26412906936612e-06,
+      "loss": 0.5285,
+      "step": 9803
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.744118183620498,
+      "learning_rate": 3.263161684083039e-06,
+      "loss": 0.5931,
+      "step": 9804
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7480824190284789,
+      "learning_rate": 3.2621943727316164e-06,
+      "loss": 0.5122,
+      "step": 9805
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6299208935569788,
+      "learning_rate": 3.261227135353025e-06,
+      "loss": 0.4177,
+      "step": 9806
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6400414694783045,
+      "learning_rate": 3.260259971988438e-06,
+      "loss": 0.4496,
+      "step": 9807
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5860832921444632,
+      "learning_rate": 3.259292882679025e-06,
+      "loss": 0.421,
+      "step": 9808
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.620047916547074,
+      "learning_rate": 3.2583258674659513e-06,
+      "loss": 0.4473,
+      "step": 9809
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5974931480253365,
+      "learning_rate": 3.2573589263903803e-06,
+      "loss": 0.4503,
+      "step": 9810
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6033952382299103,
+      "learning_rate": 3.2563920594934708e-06,
+      "loss": 0.4677,
+      "step": 9811
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7231217390976792,
+      "learning_rate": 3.255425266816379e-06,
+      "loss": 0.4535,
+      "step": 9812
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7748083867869024,
+      "learning_rate": 3.254458548400259e-06,
+      "loss": 0.5228,
+      "step": 9813
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6604880883838006,
+      "learning_rate": 3.2534919042862613e-06,
+      "loss": 0.4374,
+      "step": 9814
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7983307918453107,
+      "learning_rate": 3.2525253345155305e-06,
+      "loss": 0.5569,
+      "step": 9815
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7923631059098096,
+      "learning_rate": 3.251558839129213e-06,
+      "loss": 0.5344,
+      "step": 9816
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6550914907641446,
+      "learning_rate": 3.250592418168448e-06,
+      "loss": 0.434,
+      "step": 9817
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.576047407488201,
+      "learning_rate": 3.249626071674374e-06,
+      "loss": 0.4108,
+      "step": 9818
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6019748511060291,
+      "learning_rate": 3.2486597996881252e-06,
+      "loss": 0.479,
+      "step": 9819
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8879563645186039,
+      "learning_rate": 3.247693602250832e-06,
+      "loss": 0.5408,
+      "step": 9820
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5803954889613333,
+      "learning_rate": 3.246727479403623e-06,
+      "loss": 0.4688,
+      "step": 9821
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7346962074704275,
+      "learning_rate": 3.245761431187622e-06,
+      "loss": 0.5561,
+      "step": 9822
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.6534656074696925,
+      "learning_rate": 3.244795457643951e-06,
+      "loss": 0.5511,
+      "step": 9823
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6963013815243809,
+      "learning_rate": 3.2438295588137293e-06,
+      "loss": 0.5389,
+      "step": 9824
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7117631549899102,
+      "learning_rate": 3.2428637347380705e-06,
+      "loss": 0.5052,
+      "step": 9825
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6773070660701833,
+      "learning_rate": 3.2418979854580877e-06,
+      "loss": 0.489,
+      "step": 9826
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6941230596410103,
+      "learning_rate": 3.240932311014889e-06,
+      "loss": 0.4813,
+      "step": 9827
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6645045544506429,
+      "learning_rate": 3.2399667114495813e-06,
+      "loss": 0.467,
+      "step": 9828
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8965847741915173,
+      "learning_rate": 3.239001186803267e-06,
+      "loss": 0.5227,
+      "step": 9829
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8943685918586383,
+      "learning_rate": 3.238035737117044e-06,
+      "loss": 0.5711,
+      "step": 9830
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8220622806689967,
+      "learning_rate": 3.23707036243201e-06,
+      "loss": 0.5216,
+      "step": 9831
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.226829334532521,
+      "learning_rate": 3.2361050627892566e-06,
+      "loss": 0.5305,
+      "step": 9832
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7892365585912237,
+      "learning_rate": 3.2351398382298738e-06,
+      "loss": 0.4717,
+      "step": 9833
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6274119973413937,
+      "learning_rate": 3.234174688794949e-06,
+      "loss": 0.4727,
+      "step": 9834
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8326464561556434,
+      "learning_rate": 3.2332096145255646e-06,
+      "loss": 0.4679,
+      "step": 9835
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6593812566326863,
+      "learning_rate": 3.2322446154628013e-06,
+      "loss": 0.4642,
+      "step": 9836
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6184796831804252,
+      "learning_rate": 3.231279691647736e-06,
+      "loss": 0.4214,
+      "step": 9837
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.644655182878305,
+      "learning_rate": 3.2303148431214416e-06,
+      "loss": 0.4533,
+      "step": 9838
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6518594687625032,
+      "learning_rate": 3.229350069924989e-06,
+      "loss": 0.497,
+      "step": 9839
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6960857625066758,
+      "learning_rate": 3.2283853720994444e-06,
+      "loss": 0.4705,
+      "step": 9840
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6121445164106195,
+      "learning_rate": 3.2274207496858756e-06,
+      "loss": 0.4997,
+      "step": 9841
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7472159977548641,
+      "learning_rate": 3.2264562027253402e-06,
+      "loss": 0.4947,
+      "step": 9842
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6558324530578811,
+      "learning_rate": 3.2254917312588974e-06,
+      "loss": 0.4478,
+      "step": 9843
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7160388078735583,
+      "learning_rate": 3.224527335327601e-06,
+      "loss": 0.51,
+      "step": 9844
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7537417454398606,
+      "learning_rate": 3.223563014972502e-06,
+      "loss": 0.5289,
+      "step": 9845
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6576783491403214,
+      "learning_rate": 3.2225987702346496e-06,
+      "loss": 0.5203,
+      "step": 9846
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7367553386767534,
+      "learning_rate": 3.221634601155087e-06,
+      "loss": 0.5654,
+      "step": 9847
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8067918691951811,
+      "learning_rate": 3.2206705077748566e-06,
+      "loss": 0.5808,
+      "step": 9848
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7643784840076346,
+      "learning_rate": 3.2197064901349975e-06,
+      "loss": 0.5053,
+      "step": 9849
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7753689071548957,
+      "learning_rate": 3.2187425482765435e-06,
+      "loss": 0.4938,
+      "step": 9850
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8013427070149979,
+      "learning_rate": 3.2177786822405267e-06,
+      "loss": 0.5056,
+      "step": 9851
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6690309837945828,
+      "learning_rate": 3.216814892067975e-06,
+      "loss": 0.4412,
+      "step": 9852
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6420970393285401,
+      "learning_rate": 3.215851177799917e-06,
+      "loss": 0.4808,
+      "step": 9853
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6855795059049861,
+      "learning_rate": 3.2148875394773727e-06,
+      "loss": 0.5056,
+      "step": 9854
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.713746628553741,
+      "learning_rate": 3.2139239771413616e-06,
+      "loss": 0.4596,
+      "step": 9855
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7462333994610545,
+      "learning_rate": 3.2129604908328994e-06,
+      "loss": 0.5092,
+      "step": 9856
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7056654666830612,
+      "learning_rate": 3.2119970805929984e-06,
+      "loss": 0.4592,
+      "step": 9857
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5793630850885867,
+      "learning_rate": 3.2110337464626685e-06,
+      "loss": 0.4909,
+      "step": 9858
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.628451931771063,
+      "learning_rate": 3.2100704884829147e-06,
+      "loss": 0.4842,
+      "step": 9859
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6442462346622816,
+      "learning_rate": 3.209107306694741e-06,
+      "loss": 0.5274,
+      "step": 9860
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.6686654545925466,
+      "learning_rate": 3.2081442011391455e-06,
+      "loss": 0.5206,
+      "step": 9861
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5829889021833834,
+      "learning_rate": 3.207181171857126e-06,
+      "loss": 0.491,
+      "step": 9862
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7432435731302618,
+      "learning_rate": 3.2062182188896753e-06,
+      "loss": 0.533,
+      "step": 9863
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7329897402527397,
+      "learning_rate": 3.20525534227778e-06,
+      "loss": 0.5442,
+      "step": 9864
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7453179893366322,
+      "learning_rate": 3.204292542062432e-06,
+      "loss": 0.5029,
+      "step": 9865
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6339286054803679,
+      "learning_rate": 3.2033298182846122e-06,
+      "loss": 0.4973,
+      "step": 9866
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7891009532643618,
+      "learning_rate": 3.202367170985301e-06,
+      "loss": 0.4952,
+      "step": 9867
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5489119562476963,
+      "learning_rate": 3.201404600205475e-06,
+      "loss": 0.4355,
+      "step": 9868
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.697021613953802,
+      "learning_rate": 3.2004421059861073e-06,
+      "loss": 0.4533,
+      "step": 9869
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5738472127371834,
+      "learning_rate": 3.199479688368169e-06,
+      "loss": 0.4254,
+      "step": 9870
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7585878860833483,
+      "learning_rate": 3.1985173473926263e-06,
+      "loss": 0.5234,
+      "step": 9871
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7739078301838713,
+      "learning_rate": 3.197555083100443e-06,
+      "loss": 0.5388,
+      "step": 9872
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6693094729705964,
+      "learning_rate": 3.196592895532581e-06,
+      "loss": 0.4792,
+      "step": 9873
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5724523216615389,
+      "learning_rate": 3.195630784729996e-06,
+      "loss": 0.4429,
+      "step": 9874
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6009674516129981,
+      "learning_rate": 3.1946687507336428e-06,
+      "loss": 0.5281,
+      "step": 9875
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8148113779397542,
+      "learning_rate": 3.1937067935844718e-06,
+      "loss": 0.553,
+      "step": 9876
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7831115003604711,
+      "learning_rate": 3.1927449133234295e-06,
+      "loss": 0.5515,
+      "step": 9877
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6437030484691998,
+      "learning_rate": 3.191783109991462e-06,
+      "loss": 0.4482,
+      "step": 9878
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5867952603655965,
+      "learning_rate": 3.1908213836295083e-06,
+      "loss": 0.4757,
+      "step": 9879
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.596781784339686,
+      "learning_rate": 3.1898597342785074e-06,
+      "loss": 0.4097,
+      "step": 9880
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7115799260134911,
+      "learning_rate": 3.1888981619793946e-06,
+      "loss": 0.5045,
+      "step": 9881
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7465189151037821,
+      "learning_rate": 3.1879366667730986e-06,
+      "loss": 0.5245,
+      "step": 9882
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7501013878291433,
+      "learning_rate": 3.1869752487005485e-06,
+      "loss": 0.4648,
+      "step": 9883
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6384203260918596,
+      "learning_rate": 3.1860139078026686e-06,
+      "loss": 0.4206,
+      "step": 9884
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5984569149862544,
+      "learning_rate": 3.1850526441203794e-06,
+      "loss": 0.5041,
+      "step": 9885
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7459038381909749,
+      "learning_rate": 3.1840914576945996e-06,
+      "loss": 0.5156,
+      "step": 9886
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5813552143409968,
+      "learning_rate": 3.183130348566244e-06,
+      "loss": 0.4494,
+      "step": 9887
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7442813452056939,
+      "learning_rate": 3.1821693167762234e-06,
+      "loss": 0.4581,
+      "step": 9888
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.732492200114443,
+      "learning_rate": 3.181208362365445e-06,
+      "loss": 0.5736,
+      "step": 9889
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7917636361984168,
+      "learning_rate": 3.1802474853748157e-06,
+      "loss": 0.5086,
+      "step": 9890
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.665647202621241,
+      "learning_rate": 3.179286685845235e-06,
+      "loss": 0.4554,
+      "step": 9891
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.563271799879308,
+      "learning_rate": 3.178325963817603e-06,
+      "loss": 0.4229,
+      "step": 9892
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7276607686629478,
+      "learning_rate": 3.1773653193328136e-06,
+      "loss": 0.467,
+      "step": 9893
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6964931286636661,
+      "learning_rate": 3.1764047524317578e-06,
+      "loss": 0.4941,
+      "step": 9894
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7717033031170665,
+      "learning_rate": 3.175444263155324e-06,
+      "loss": 0.5581,
+      "step": 9895
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5420426352527334,
+      "learning_rate": 3.1744838515443976e-06,
+      "loss": 0.4318,
+      "step": 9896
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7670061772366024,
+      "learning_rate": 3.1735235176398604e-06,
+      "loss": 0.4788,
+      "step": 9897
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7983310279508355,
+      "learning_rate": 3.17256326148259e-06,
+      "loss": 0.5207,
+      "step": 9898
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8397255911643894,
+      "learning_rate": 3.1716030831134627e-06,
+      "loss": 0.4903,
+      "step": 9899
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7256707202020258,
+      "learning_rate": 3.1706429825733486e-06,
+      "loss": 0.4844,
+      "step": 9900
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5689093431868085,
+      "learning_rate": 3.1696829599031176e-06,
+      "loss": 0.4266,
+      "step": 9901
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6262273340470185,
+      "learning_rate": 3.1687230151436322e-06,
+      "loss": 0.4583,
+      "step": 9902
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.670657683213048,
+      "learning_rate": 3.1677631483357572e-06,
+      "loss": 0.4738,
+      "step": 9903
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6337897363984735,
+      "learning_rate": 3.1668033595203505e-06,
+      "loss": 0.4749,
+      "step": 9904
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6891063871269649,
+      "learning_rate": 3.165843648738267e-06,
+      "loss": 0.4953,
+      "step": 9905
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5968432218174194,
+      "learning_rate": 3.1648840160303585e-06,
+      "loss": 0.4714,
+      "step": 9906
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8001986283879055,
+      "learning_rate": 3.163924461437472e-06,
+      "loss": 0.5573,
+      "step": 9907
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7215817195717713,
+      "learning_rate": 3.1629649850004553e-06,
+      "loss": 0.5606,
+      "step": 9908
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6206732567557802,
+      "learning_rate": 3.1620055867601487e-06,
+      "loss": 0.4517,
+      "step": 9909
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7091691165945644,
+      "learning_rate": 3.16104626675739e-06,
+      "loss": 0.5157,
+      "step": 9910
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7731843108203841,
+      "learning_rate": 3.1600870250330155e-06,
+      "loss": 0.5627,
+      "step": 9911
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8672547607346605,
+      "learning_rate": 3.1591278616278574e-06,
+      "loss": 0.5336,
+      "step": 9912
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.795869415216268,
+      "learning_rate": 3.158168776582743e-06,
+      "loss": 0.5074,
+      "step": 9913
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6715420976532219,
+      "learning_rate": 3.157209769938496e-06,
+      "loss": 0.4916,
+      "step": 9914
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.607640260436024,
+      "learning_rate": 3.1562508417359433e-06,
+      "loss": 0.462,
+      "step": 9915
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7417548998133111,
+      "learning_rate": 3.1552919920159e-06,
+      "loss": 0.5082,
+      "step": 9916
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6717004473914451,
+      "learning_rate": 3.154333220819182e-06,
+      "loss": 0.5019,
+      "step": 9917
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6181879367924511,
+      "learning_rate": 3.1533745281866e-06,
+      "loss": 0.4976,
+      "step": 9918
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7119358765643872,
+      "learning_rate": 3.1524159141589644e-06,
+      "loss": 0.4902,
+      "step": 9919
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7530063034176033,
+      "learning_rate": 3.1514573787770786e-06,
+      "loss": 0.5051,
+      "step": 9920
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7154305445386362,
+      "learning_rate": 3.1504989220817457e-06,
+      "loss": 0.5079,
+      "step": 9921
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7660282096299629,
+      "learning_rate": 3.1495405441137627e-06,
+      "loss": 0.5076,
+      "step": 9922
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6026131679881637,
+      "learning_rate": 3.1485822449139263e-06,
+      "loss": 0.4755,
+      "step": 9923
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7836022467326218,
+      "learning_rate": 3.1476240245230272e-06,
+      "loss": 0.5113,
+      "step": 9924
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6876687248022918,
+      "learning_rate": 3.1466658829818543e-06,
+      "loss": 0.4817,
+      "step": 9925
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6415848110272832,
+      "learning_rate": 3.1457078203311897e-06,
+      "loss": 0.4566,
+      "step": 9926
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7808714798928587,
+      "learning_rate": 3.14474983661182e-06,
+      "loss": 0.5687,
+      "step": 9927
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7250699538423234,
+      "learning_rate": 3.1437919318645216e-06,
+      "loss": 0.5137,
+      "step": 9928
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7075694266545647,
+      "learning_rate": 3.1428341061300687e-06,
+      "loss": 0.4797,
+      "step": 9929
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5473386506540765,
+      "learning_rate": 3.141876359449233e-06,
+      "loss": 0.4402,
+      "step": 9930
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7825510789290535,
+      "learning_rate": 3.140918691862783e-06,
+      "loss": 0.5629,
+      "step": 9931
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.880621228415058,
+      "learning_rate": 3.139961103411484e-06,
+      "loss": 0.4939,
+      "step": 9932
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.1042619234925284,
+      "learning_rate": 3.1390035941360975e-06,
+      "loss": 0.5651,
+      "step": 9933
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6925194000489686,
+      "learning_rate": 3.1380461640773803e-06,
+      "loss": 0.5025,
+      "step": 9934
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6254911281763132,
+      "learning_rate": 3.137088813276088e-06,
+      "loss": 0.4415,
+      "step": 9935
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6445112189690168,
+      "learning_rate": 3.136131541772972e-06,
+      "loss": 0.4999,
+      "step": 9936
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5818518250465591,
+      "learning_rate": 3.13517434960878e-06,
+      "loss": 0.4901,
+      "step": 9937
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7642835612508834,
+      "learning_rate": 3.134217236824257e-06,
+      "loss": 0.5598,
+      "step": 9938
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.663430748715996,
+      "learning_rate": 3.133260203460143e-06,
+      "loss": 0.5314,
+      "step": 9939
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.623174695249876,
+      "learning_rate": 3.1323032495571777e-06,
+      "loss": 0.4703,
+      "step": 9940
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6523327564812906,
+      "learning_rate": 3.1313463751560935e-06,
+      "loss": 0.4644,
+      "step": 9941
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.0993333002394308,
+      "learning_rate": 3.1303895802976245e-06,
+      "loss": 0.5285,
+      "step": 9942
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6651099934805494,
+      "learning_rate": 3.1294328650224965e-06,
+      "loss": 0.506,
+      "step": 9943
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6657817185498772,
+      "learning_rate": 3.128476229371433e-06,
+      "loss": 0.4638,
+      "step": 9944
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6483782927121519,
+      "learning_rate": 3.1275196733851575e-06,
+      "loss": 0.4324,
+      "step": 9945
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7462880908031876,
+      "learning_rate": 3.1265631971043854e-06,
+      "loss": 0.5101,
+      "step": 9946
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7695180921923955,
+      "learning_rate": 3.1256068005698303e-06,
+      "loss": 0.4851,
+      "step": 9947
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5995232463677926,
+      "learning_rate": 3.1246504838222047e-06,
+      "loss": 0.447,
+      "step": 9948
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7071339803035661,
+      "learning_rate": 3.123694246902216e-06,
+      "loss": 0.4382,
+      "step": 9949
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8026421358645626,
+      "learning_rate": 3.1227380898505665e-06,
+      "loss": 0.499,
+      "step": 9950
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5760689060702163,
+      "learning_rate": 3.121782012707957e-06,
+      "loss": 0.4422,
+      "step": 9951
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6093306136261478,
+      "learning_rate": 3.120826015515086e-06,
+      "loss": 0.4871,
+      "step": 9952
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7926944690816904,
+      "learning_rate": 3.1198700983126464e-06,
+      "loss": 0.5273,
+      "step": 9953
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6148644236290115,
+      "learning_rate": 3.1189142611413286e-06,
+      "loss": 0.469,
+      "step": 9954
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.827937456644679,
+      "learning_rate": 3.1179585040418204e-06,
+      "loss": 0.4898,
+      "step": 9955
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.60952679725037,
+      "learning_rate": 3.1170028270548036e-06,
+      "loss": 0.44,
+      "step": 9956
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6197475712043369,
+      "learning_rate": 3.11604723022096e-06,
+      "loss": 0.4278,
+      "step": 9957
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5446321286627389,
+      "learning_rate": 3.1150917135809654e-06,
+      "loss": 0.4496,
+      "step": 9958
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9763700108494906,
+      "learning_rate": 3.114136277175493e-06,
+      "loss": 0.4882,
+      "step": 9959
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5451786417747946,
+      "learning_rate": 3.113180921045213e-06,
+      "loss": 0.4078,
+      "step": 9960
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5711121002153202,
+      "learning_rate": 3.112225645230792e-06,
+      "loss": 0.4433,
+      "step": 9961
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6564735920284965,
+      "learning_rate": 3.111270449772892e-06,
+      "loss": 0.4842,
+      "step": 9962
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.720590846859446,
+      "learning_rate": 3.1103153347121743e-06,
+      "loss": 0.5187,
+      "step": 9963
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6874548830380423,
+      "learning_rate": 3.1093603000892923e-06,
+      "loss": 0.5016,
+      "step": 9964
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7900460478602355,
+      "learning_rate": 3.1084053459449025e-06,
+      "loss": 0.5182,
+      "step": 9965
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7711855452182235,
+      "learning_rate": 3.107450472319652e-06,
+      "loss": 0.5315,
+      "step": 9966
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.612830530676528,
+      "learning_rate": 3.1064956792541877e-06,
+      "loss": 0.4986,
+      "step": 9967
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.574623794892652,
+      "learning_rate": 3.1055409667891513e-06,
+      "loss": 0.4765,
+      "step": 9968
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6902338288936075,
+      "learning_rate": 3.1045863349651827e-06,
+      "loss": 0.4469,
+      "step": 9969
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5685269148314458,
+      "learning_rate": 3.1036317838229162e-06,
+      "loss": 0.4537,
+      "step": 9970
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5456361567034992,
+      "learning_rate": 3.1026773134029854e-06,
+      "loss": 0.4541,
+      "step": 9971
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6217646636411817,
+      "learning_rate": 3.1017229237460176e-06,
+      "loss": 0.4837,
+      "step": 9972
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.2924020366665425,
+      "learning_rate": 3.1007686148926396e-06,
+      "loss": 0.4837,
+      "step": 9973
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6089581590918436,
+      "learning_rate": 3.0998143868834717e-06,
+      "loss": 0.4703,
+      "step": 9974
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7653251331614317,
+      "learning_rate": 3.0988602397591338e-06,
+      "loss": 0.482,
+      "step": 9975
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5371536887762518,
+      "learning_rate": 3.097906173560239e-06,
+      "loss": 0.4225,
+      "step": 9976
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7761111294942714,
+      "learning_rate": 3.096952188327401e-06,
+      "loss": 0.4499,
+      "step": 9977
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6802528396942026,
+      "learning_rate": 3.0959982841012283e-06,
+      "loss": 0.4741,
+      "step": 9978
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6928241219339091,
+      "learning_rate": 3.095044460922323e-06,
+      "loss": 0.4772,
+      "step": 9979
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7108420253509063,
+      "learning_rate": 3.0940907188312885e-06,
+      "loss": 0.4493,
+      "step": 9980
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.602090956177948,
+      "learning_rate": 3.0931370578687214e-06,
+      "loss": 0.4959,
+      "step": 9981
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6395492350721127,
+      "learning_rate": 3.0921834780752163e-06,
+      "loss": 0.5153,
+      "step": 9982
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.601243451801059,
+      "learning_rate": 3.0912299794913636e-06,
+      "loss": 0.5149,
+      "step": 9983
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.956703778109413,
+      "learning_rate": 3.090276562157752e-06,
+      "loss": 0.5875,
+      "step": 9984
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7156525862103044,
+      "learning_rate": 3.089323226114964e-06,
+      "loss": 0.5059,
+      "step": 9985
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6957222003615686,
+      "learning_rate": 3.0883699714035807e-06,
+      "loss": 0.4119,
+      "step": 9986
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.583525071761603,
+      "learning_rate": 3.087416798064179e-06,
+      "loss": 0.4592,
+      "step": 9987
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6550509721118933,
+      "learning_rate": 3.0864637061373306e-06,
+      "loss": 0.4695,
+      "step": 9988
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6899522372675596,
+      "learning_rate": 3.0855106956636095e-06,
+      "loss": 0.5119,
+      "step": 9989
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6310101863850072,
+      "learning_rate": 3.08455776668358e-06,
+      "loss": 0.4036,
+      "step": 9990
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6086763255538692,
+      "learning_rate": 3.083604919237806e-06,
+      "loss": 0.4114,
+      "step": 9991
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6767013629384223,
+      "learning_rate": 3.082652153366846e-06,
+      "loss": 0.4504,
+      "step": 9992
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6882547989452196,
+      "learning_rate": 3.0816994691112567e-06,
+      "loss": 0.4779,
+      "step": 9993
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7646097099019413,
+      "learning_rate": 3.0807468665115913e-06,
+      "loss": 0.5501,
+      "step": 9994
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5742321009665137,
+      "learning_rate": 3.079794345608399e-06,
+      "loss": 0.4549,
+      "step": 9995
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5053944630526432,
+      "learning_rate": 3.0788419064422256e-06,
+      "loss": 0.3972,
+      "step": 9996
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5703044952933406,
+      "learning_rate": 3.0778895490536124e-06,
+      "loss": 0.4103,
+      "step": 9997
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7749884170205388,
+      "learning_rate": 3.0769372734830994e-06,
+      "loss": 0.5116,
+      "step": 9998
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8219816101359045,
+      "learning_rate": 3.075985079771221e-06,
+      "loss": 0.5343,
+      "step": 9999
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7689351467617338,
+      "learning_rate": 3.07503296795851e-06,
+      "loss": 0.5247,
+      "step": 10000
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7387239325911442,
+      "learning_rate": 3.074080938085493e-06,
+      "loss": 0.4952,
+      "step": 10001
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.0615549005643687,
+      "learning_rate": 3.0731289901926964e-06,
+      "loss": 0.5688,
+      "step": 10002
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7533877985632432,
+      "learning_rate": 3.072177124320641e-06,
+      "loss": 0.5156,
+      "step": 10003
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7618768628850636,
+      "learning_rate": 3.0712253405098456e-06,
+      "loss": 0.5744,
+      "step": 10004
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7585952723183147,
+      "learning_rate": 3.0702736388008247e-06,
+      "loss": 0.5398,
+      "step": 10005
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7553428966500044,
+      "learning_rate": 3.0693220192340876e-06,
+      "loss": 0.5206,
+      "step": 10006
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6108437038033367,
+      "learning_rate": 3.0683704818501438e-06,
+      "loss": 0.48,
+      "step": 10007
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8480712547236272,
+      "learning_rate": 3.067419026689495e-06,
+      "loss": 0.5152,
+      "step": 10008
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7075026472470567,
+      "learning_rate": 3.066467653792643e-06,
+      "loss": 0.4912,
+      "step": 10009
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7245669215115459,
+      "learning_rate": 3.065516363200084e-06,
+      "loss": 0.4735,
+      "step": 10010
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7473986085675763,
+      "learning_rate": 3.0645651549523114e-06,
+      "loss": 0.5026,
+      "step": 10011
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5899756990735727,
+      "learning_rate": 3.0636140290898166e-06,
+      "loss": 0.4779,
+      "step": 10012
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8824406665943079,
+      "learning_rate": 3.062662985653082e-06,
+      "loss": 0.518,
+      "step": 10013
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.713828159259407,
+      "learning_rate": 3.061712024682596e-06,
+      "loss": 0.5213,
+      "step": 10014
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7137738184450169,
+      "learning_rate": 3.0607611462188342e-06,
+      "loss": 0.4667,
+      "step": 10015
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6984211698335809,
+      "learning_rate": 3.0598103503022734e-06,
+      "loss": 0.4949,
+      "step": 10016
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5803382187767104,
+      "learning_rate": 3.0588596369733863e-06,
+      "loss": 0.4371,
+      "step": 10017
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.704751079108312,
+      "learning_rate": 3.0579090062726415e-06,
+      "loss": 0.5002,
+      "step": 10018
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7865243535100296,
+      "learning_rate": 3.0569584582405044e-06,
+      "loss": 0.5106,
+      "step": 10019
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6013805575762681,
+      "learning_rate": 3.056007992917437e-06,
+      "loss": 0.4607,
+      "step": 10020
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6338076601598049,
+      "learning_rate": 3.0550576103438966e-06,
+      "loss": 0.4666,
+      "step": 10021
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8577282948578187,
+      "learning_rate": 3.0541073105603387e-06,
+      "loss": 0.5421,
+      "step": 10022
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7320387177694757,
+      "learning_rate": 3.053157093607214e-06,
+      "loss": 0.5511,
+      "step": 10023
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6883125579858914,
+      "learning_rate": 3.0522069595249717e-06,
+      "loss": 0.4824,
+      "step": 10024
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6183742614961625,
+      "learning_rate": 3.0512569083540537e-06,
+      "loss": 0.4445,
+      "step": 10025
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7261391838529112,
+      "learning_rate": 3.0503069401349006e-06,
+      "loss": 0.5036,
+      "step": 10026
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.651606948081618,
+      "learning_rate": 3.0493570549079522e-06,
+      "loss": 0.4829,
+      "step": 10027
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.676641055208218,
+      "learning_rate": 3.048407252713641e-06,
+      "loss": 0.513,
+      "step": 10028
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7831625782380829,
+      "learning_rate": 3.0474575335923967e-06,
+      "loss": 0.5288,
+      "step": 10029
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7136661129400382,
+      "learning_rate": 3.046507897584645e-06,
+      "loss": 0.5211,
+      "step": 10030
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8144461603238857,
+      "learning_rate": 3.0455583447308105e-06,
+      "loss": 0.4883,
+      "step": 10031
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7931373990728641,
+      "learning_rate": 3.0446088750713123e-06,
+      "loss": 0.5091,
+      "step": 10032
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5697133546601547,
+      "learning_rate": 3.043659488646564e-06,
+      "loss": 0.4389,
+      "step": 10033
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5604022302195278,
+      "learning_rate": 3.042710185496981e-06,
+      "loss": 0.437,
+      "step": 10034
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.641352831044806,
+      "learning_rate": 3.041760965662971e-06,
+      "loss": 0.4454,
+      "step": 10035
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8715159717674851,
+      "learning_rate": 3.040811829184938e-06,
+      "loss": 0.5269,
+      "step": 10036
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8903074305787414,
+      "learning_rate": 3.039862776103285e-06,
+      "loss": 0.4897,
+      "step": 10037
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.9696778903792544,
+      "learning_rate": 3.0389138064584084e-06,
+      "loss": 0.5127,
+      "step": 10038
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6020712900758779,
+      "learning_rate": 3.037964920290706e-06,
+      "loss": 0.4542,
+      "step": 10039
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7057177378866583,
+      "learning_rate": 3.0370161176405676e-06,
+      "loss": 0.4966,
+      "step": 10040
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8297266839166492,
+      "learning_rate": 3.0360673985483803e-06,
+      "loss": 0.5246,
+      "step": 10041
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5756645173131348,
+      "learning_rate": 3.0351187630545277e-06,
+      "loss": 0.4836,
+      "step": 10042
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7288508495495228,
+      "learning_rate": 3.03417021119939e-06,
+      "loss": 0.5216,
+      "step": 10043
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6407747002090415,
+      "learning_rate": 3.0332217430233453e-06,
+      "loss": 0.4732,
+      "step": 10044
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8804538125993043,
+      "learning_rate": 3.0322733585667653e-06,
+      "loss": 0.4997,
+      "step": 10045
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6284289269867019,
+      "learning_rate": 3.0313250578700217e-06,
+      "loss": 0.4698,
+      "step": 10046
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7854980850057762,
+      "learning_rate": 3.030376840973478e-06,
+      "loss": 0.5258,
+      "step": 10047
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7068188670250871,
+      "learning_rate": 3.0294287079174987e-06,
+      "loss": 0.5143,
+      "step": 10048
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7952001354525379,
+      "learning_rate": 3.0284806587424424e-06,
+      "loss": 0.5979,
+      "step": 10049
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7092844791672158,
+      "learning_rate": 3.0275326934886628e-06,
+      "loss": 0.4552,
+      "step": 10050
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6795123172734648,
+      "learning_rate": 3.0265848121965145e-06,
+      "loss": 0.507,
+      "step": 10051
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6838918044613908,
+      "learning_rate": 3.0256370149063453e-06,
+      "loss": 0.5623,
+      "step": 10052
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8283735266829618,
+      "learning_rate": 3.0246893016584987e-06,
+      "loss": 0.525,
+      "step": 10053
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8431487495445705,
+      "learning_rate": 3.0237416724933166e-06,
+      "loss": 0.5078,
+      "step": 10054
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.618514115349961,
+      "learning_rate": 3.022794127451136e-06,
+      "loss": 0.4861,
+      "step": 10055
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7774048537524458,
+      "learning_rate": 3.021846666572291e-06,
+      "loss": 0.507,
+      "step": 10056
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6343040788754329,
+      "learning_rate": 3.020899289897113e-06,
+      "loss": 0.4459,
+      "step": 10057
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6149030932862146,
+      "learning_rate": 3.0199519974659276e-06,
+      "loss": 0.4797,
+      "step": 10058
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8142902832891326,
+      "learning_rate": 3.0190047893190575e-06,
+      "loss": 0.5064,
+      "step": 10059
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5875697442357543,
+      "learning_rate": 3.0180576654968242e-06,
+      "loss": 0.4749,
+      "step": 10060
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5807271333060996,
+      "learning_rate": 3.017110626039542e-06,
+      "loss": 0.4328,
+      "step": 10061
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8156702292407569,
+      "learning_rate": 3.0161636709875245e-06,
+      "loss": 0.4791,
+      "step": 10062
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8883279805108075,
+      "learning_rate": 3.0152168003810788e-06,
+      "loss": 0.5289,
+      "step": 10063
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7966769391439963,
+      "learning_rate": 3.014270014260513e-06,
+      "loss": 0.5244,
+      "step": 10064
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.303583516155905,
+      "learning_rate": 3.013323312666126e-06,
+      "loss": 0.4876,
+      "step": 10065
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.58675524657906,
+      "learning_rate": 3.0123766956382177e-06,
+      "loss": 0.4469,
+      "step": 10066
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7589867412843092,
+      "learning_rate": 3.011430163217082e-06,
+      "loss": 0.586,
+      "step": 10067
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7693628082801335,
+      "learning_rate": 3.01048371544301e-06,
+      "loss": 0.5209,
+      "step": 10068
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5944715712428372,
+      "learning_rate": 3.009537352356289e-06,
+      "loss": 0.4545,
+      "step": 10069
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7005635876539956,
+      "learning_rate": 3.0085910739972013e-06,
+      "loss": 0.4282,
+      "step": 10070
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8225769212919801,
+      "learning_rate": 3.0076448804060286e-06,
+      "loss": 0.5229,
+      "step": 10071
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.0036227321852822,
+      "learning_rate": 3.006698771623046e-06,
+      "loss": 0.472,
+      "step": 10072
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5925436969464989,
+      "learning_rate": 3.005752747688528e-06,
+      "loss": 0.44,
+      "step": 10073
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6921521343201524,
+      "learning_rate": 3.004806808642742e-06,
+      "loss": 0.463,
+      "step": 10074
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.60239773074461,
+      "learning_rate": 3.003860954525954e-06,
+      "loss": 0.4626,
+      "step": 10075
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7077316777280128,
+      "learning_rate": 3.002915185378427e-06,
+      "loss": 0.5023,
+      "step": 10076
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7522666144284539,
+      "learning_rate": 3.0019695012404193e-06,
+      "loss": 0.5559,
+      "step": 10077
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8185355937073601,
+      "learning_rate": 3.001023902152185e-06,
+      "loss": 0.5093,
+      "step": 10078
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.9198465109077911,
+      "learning_rate": 3.0000783881539747e-06,
+      "loss": 0.5868,
+      "step": 10079
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8550110783014858,
+      "learning_rate": 2.9991329592860376e-06,
+      "loss": 0.5043,
+      "step": 10080
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.734583464318067,
+      "learning_rate": 2.998187615588616e-06,
+      "loss": 0.4729,
+      "step": 10081
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6177460097427718,
+      "learning_rate": 2.9972423571019503e-06,
+      "loss": 0.4996,
+      "step": 10082
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.9354631800349099,
+      "learning_rate": 2.996297183866278e-06,
+      "loss": 0.5243,
+      "step": 10083
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.9388552029761114,
+      "learning_rate": 2.995352095921832e-06,
+      "loss": 0.4607,
+      "step": 10084
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.673897873608996,
+      "learning_rate": 2.9944070933088408e-06,
+      "loss": 0.5089,
+      "step": 10085
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7970155534274046,
+      "learning_rate": 2.9934621760675308e-06,
+      "loss": 0.5325,
+      "step": 10086
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6616889414459125,
+      "learning_rate": 2.992517344238124e-06,
+      "loss": 0.4741,
+      "step": 10087
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6796734364489979,
+      "learning_rate": 2.991572597860837e-06,
+      "loss": 0.4344,
+      "step": 10088
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7216796163782366,
+      "learning_rate": 2.990627936975888e-06,
+      "loss": 0.4914,
+      "step": 10089
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7493563605490247,
+      "learning_rate": 2.989683361623488e-06,
+      "loss": 0.4512,
+      "step": 10090
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6755188370213615,
+      "learning_rate": 2.988738871843842e-06,
+      "loss": 0.4529,
+      "step": 10091
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9182093900054465,
+      "learning_rate": 2.9877944676771554e-06,
+      "loss": 0.5545,
+      "step": 10092
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.1928959872543183,
+      "learning_rate": 2.986850149163628e-06,
+      "loss": 0.5176,
+      "step": 10093
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.737471400068593,
+      "learning_rate": 2.985905916343458e-06,
+      "loss": 0.5067,
+      "step": 10094
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6022972409017776,
+      "learning_rate": 2.9849617692568356e-06,
+      "loss": 0.4609,
+      "step": 10095
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8303501827301287,
+      "learning_rate": 2.984017707943952e-06,
+      "loss": 0.5094,
+      "step": 10096
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.876549585807127,
+      "learning_rate": 2.983073732444992e-06,
+      "loss": 0.5175,
+      "step": 10097
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.594992338486311,
+      "learning_rate": 2.982129842800139e-06,
+      "loss": 0.4495,
+      "step": 10098
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7201306166707563,
+      "learning_rate": 2.9811860390495694e-06,
+      "loss": 0.4328,
+      "step": 10099
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8887686741218395,
+      "learning_rate": 2.9802423212334575e-06,
+      "loss": 0.4785,
+      "step": 10100
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8867668237412709,
+      "learning_rate": 2.9792986893919774e-06,
+      "loss": 0.5371,
+      "step": 10101
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9830089174735464,
+      "learning_rate": 2.9783551435652947e-06,
+      "loss": 0.467,
+      "step": 10102
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5411161149393208,
+      "learning_rate": 2.977411683793574e-06,
+      "loss": 0.4248,
+      "step": 10103
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6652680659069923,
+      "learning_rate": 2.9764683101169746e-06,
+      "loss": 0.4859,
+      "step": 10104
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8077801746704958,
+      "learning_rate": 2.9755250225756525e-06,
+      "loss": 0.5204,
+      "step": 10105
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7764389497501268,
+      "learning_rate": 2.974581821209761e-06,
+      "loss": 0.5202,
+      "step": 10106
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7197404009164483,
+      "learning_rate": 2.9736387060594484e-06,
+      "loss": 0.4562,
+      "step": 10107
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8357037768159049,
+      "learning_rate": 2.9726956771648616e-06,
+      "loss": 0.5501,
+      "step": 10108
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.67724049421269,
+      "learning_rate": 2.971752734566142e-06,
+      "loss": 0.5183,
+      "step": 10109
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7044068443668652,
+      "learning_rate": 2.9708098783034263e-06,
+      "loss": 0.4752,
+      "step": 10110
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7033922958840514,
+      "learning_rate": 2.9698671084168497e-06,
+      "loss": 0.5317,
+      "step": 10111
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7709944479576237,
+      "learning_rate": 2.968924424946541e-06,
+      "loss": 0.5065,
+      "step": 10112
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6444397679009273,
+      "learning_rate": 2.967981827932631e-06,
+      "loss": 0.4504,
+      "step": 10113
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6266647373921675,
+      "learning_rate": 2.9670393174152417e-06,
+      "loss": 0.4897,
+      "step": 10114
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7758081475425422,
+      "learning_rate": 2.966096893434492e-06,
+      "loss": 0.5538,
+      "step": 10115
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.639264055117079,
+      "learning_rate": 2.9651545560304986e-06,
+      "loss": 0.4487,
+      "step": 10116
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7670169816847134,
+      "learning_rate": 2.9642123052433725e-06,
+      "loss": 0.5225,
+      "step": 10117
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.765958751322079,
+      "learning_rate": 2.963270141113223e-06,
+      "loss": 0.5261,
+      "step": 10118
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7417938385491667,
+      "learning_rate": 2.9623280636801554e-06,
+      "loss": 0.4577,
+      "step": 10119
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6330606280845634,
+      "learning_rate": 2.9613860729842714e-06,
+      "loss": 0.4686,
+      "step": 10120
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8042711924059736,
+      "learning_rate": 2.960444169065667e-06,
+      "loss": 0.5317,
+      "step": 10121
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6931330767737331,
+      "learning_rate": 2.959502351964437e-06,
+      "loss": 0.4783,
+      "step": 10122
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6227064661592011,
+      "learning_rate": 2.9585606217206714e-06,
+      "loss": 0.4434,
+      "step": 10123
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6326883520728792,
+      "learning_rate": 2.957618978374457e-06,
+      "loss": 0.4877,
+      "step": 10124
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7003053110487208,
+      "learning_rate": 2.9566774219658744e-06,
+      "loss": 0.447,
+      "step": 10125
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6386031879147133,
+      "learning_rate": 2.955735952535006e-06,
+      "loss": 0.4468,
+      "step": 10126
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7669863806583794,
+      "learning_rate": 2.9547945701219245e-06,
+      "loss": 0.4719,
+      "step": 10127
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6717987016775734,
+      "learning_rate": 2.9538532747667036e-06,
+      "loss": 0.4692,
+      "step": 10128
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8611211247390582,
+      "learning_rate": 2.9529120665094103e-06,
+      "loss": 0.5389,
+      "step": 10129
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5387871274834539,
+      "learning_rate": 2.951970945390108e-06,
+      "loss": 0.4386,
+      "step": 10130
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6376877981854637,
+      "learning_rate": 2.9510299114488592e-06,
+      "loss": 0.4805,
+      "step": 10131
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6348081336631767,
+      "learning_rate": 2.9500889647257193e-06,
+      "loss": 0.5089,
+      "step": 10132
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.645884238847916,
+      "learning_rate": 2.9491481052607406e-06,
+      "loss": 0.5079,
+      "step": 10133
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6452682051769816,
+      "learning_rate": 2.9482073330939743e-06,
+      "loss": 0.4876,
+      "step": 10134
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8376356155256386,
+      "learning_rate": 2.947266648265465e-06,
+      "loss": 0.5016,
+      "step": 10135
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6292333787968463,
+      "learning_rate": 2.9463260508152547e-06,
+      "loss": 0.4734,
+      "step": 10136
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8373799011921564,
+      "learning_rate": 2.945385540783381e-06,
+      "loss": 0.5214,
+      "step": 10137
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6615495878018989,
+      "learning_rate": 2.9444451182098805e-06,
+      "loss": 0.5056,
+      "step": 10138
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.201925206034663,
+      "learning_rate": 2.9435047831347824e-06,
+      "loss": 0.5585,
+      "step": 10139
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.1188336960899259,
+      "learning_rate": 2.9425645355981143e-06,
+      "loss": 0.5157,
+      "step": 10140
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7477213854610992,
+      "learning_rate": 2.9416243756398986e-06,
+      "loss": 0.4936,
+      "step": 10141
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7815082644484085,
+      "learning_rate": 2.9406843033001563e-06,
+      "loss": 0.4832,
+      "step": 10142
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8209542514004152,
+      "learning_rate": 2.9397443186189024e-06,
+      "loss": 0.5489,
+      "step": 10143
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6637044558774606,
+      "learning_rate": 2.9388044216361492e-06,
+      "loss": 0.4937,
+      "step": 10144
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5880107320278892,
+      "learning_rate": 2.9378646123919054e-06,
+      "loss": 0.4471,
+      "step": 10145
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8429328183645565,
+      "learning_rate": 2.936924890926175e-06,
+      "loss": 0.5619,
+      "step": 10146
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7516802532245696,
+      "learning_rate": 2.9359852572789594e-06,
+      "loss": 0.5099,
+      "step": 10147
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6508521883211092,
+      "learning_rate": 2.9350457114902565e-06,
+      "loss": 0.479,
+      "step": 10148
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6544020286912982,
+      "learning_rate": 2.934106253600058e-06,
+      "loss": 0.4783,
+      "step": 10149
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6214310053089499,
+      "learning_rate": 2.933166883648353e-06,
+      "loss": 0.5004,
+      "step": 10150
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.2185275914777103,
+      "learning_rate": 2.932227601675131e-06,
+      "loss": 0.5462,
+      "step": 10151
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7644958094851502,
+      "learning_rate": 2.9312884077203725e-06,
+      "loss": 0.4934,
+      "step": 10152
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5779097297825884,
+      "learning_rate": 2.930349301824056e-06,
+      "loss": 0.4473,
+      "step": 10153
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6559143254130309,
+      "learning_rate": 2.929410284026156e-06,
+      "loss": 0.5137,
+      "step": 10154
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.725209318894737,
+      "learning_rate": 2.9284713543666434e-06,
+      "loss": 0.528,
+      "step": 10155
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7034797640431257,
+      "learning_rate": 2.9275325128854854e-06,
+      "loss": 0.5068,
+      "step": 10156
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5858274836647018,
+      "learning_rate": 2.926593759622646e-06,
+      "loss": 0.4194,
+      "step": 10157
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6500672658007484,
+      "learning_rate": 2.925655094618084e-06,
+      "loss": 0.4793,
+      "step": 10158
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7217885520479481,
+      "learning_rate": 2.9247165179117567e-06,
+      "loss": 0.4786,
+      "step": 10159
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.634553578264273,
+      "learning_rate": 2.9237780295436153e-06,
+      "loss": 0.4835,
+      "step": 10160
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8408847210221582,
+      "learning_rate": 2.9228396295536083e-06,
+      "loss": 0.5354,
+      "step": 10161
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7644156960060315,
+      "learning_rate": 2.921901317981679e-06,
+      "loss": 0.5325,
+      "step": 10162
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8285579352659967,
+      "learning_rate": 2.9209630948677715e-06,
+      "loss": 0.5264,
+      "step": 10163
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7487038187038251,
+      "learning_rate": 2.9200249602518215e-06,
+      "loss": 0.5284,
+      "step": 10164
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6425912550011786,
+      "learning_rate": 2.9190869141737628e-06,
+      "loss": 0.5401,
+      "step": 10165
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7359199837551312,
+      "learning_rate": 2.9181489566735256e-06,
+      "loss": 0.5379,
+      "step": 10166
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8129476332300664,
+      "learning_rate": 2.917211087791032e-06,
+      "loss": 0.5472,
+      "step": 10167
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6923659456726867,
+      "learning_rate": 2.9162733075662087e-06,
+      "loss": 0.4581,
+      "step": 10168
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7804021163281288,
+      "learning_rate": 2.91533561603897e-06,
+      "loss": 0.5254,
+      "step": 10169
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8436480174237487,
+      "learning_rate": 2.9143980132492334e-06,
+      "loss": 0.5892,
+      "step": 10170
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8545459823577821,
+      "learning_rate": 2.9134604992369097e-06,
+      "loss": 0.536,
+      "step": 10171
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.654194378378166,
+      "learning_rate": 2.9125230740419037e-06,
+      "loss": 0.4624,
+      "step": 10172
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6705653229842632,
+      "learning_rate": 2.911585737704121e-06,
+      "loss": 0.4559,
+      "step": 10173
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6687289069006598,
+      "learning_rate": 2.910648490263459e-06,
+      "loss": 0.4759,
+      "step": 10174
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6442381443013449,
+      "learning_rate": 2.9097113317598135e-06,
+      "loss": 0.4706,
+      "step": 10175
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6667896788204748,
+      "learning_rate": 2.908774262233079e-06,
+      "loss": 0.4529,
+      "step": 10176
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5904164099469337,
+      "learning_rate": 2.907837281723139e-06,
+      "loss": 0.4583,
+      "step": 10177
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.55932948275627,
+      "learning_rate": 2.9069003902698833e-06,
+      "loss": 0.4731,
+      "step": 10178
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8565273952319515,
+      "learning_rate": 2.905963587913187e-06,
+      "loss": 0.4672,
+      "step": 10179
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7750285684129028,
+      "learning_rate": 2.905026874692931e-06,
+      "loss": 0.5748,
+      "step": 10180
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7166605566136042,
+      "learning_rate": 2.904090250648985e-06,
+      "loss": 0.5358,
+      "step": 10181
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6339450470409902,
+      "learning_rate": 2.903153715821221e-06,
+      "loss": 0.4702,
+      "step": 10182
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.9731304868139452,
+      "learning_rate": 2.9022172702495004e-06,
+      "loss": 0.5219,
+      "step": 10183
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7514882638573377,
+      "learning_rate": 2.9012809139736895e-06,
+      "loss": 0.4829,
+      "step": 10184
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7475335600527702,
+      "learning_rate": 2.9003446470336415e-06,
+      "loss": 0.5333,
+      "step": 10185
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6985193720968823,
+      "learning_rate": 2.899408469469214e-06,
+      "loss": 0.5511,
+      "step": 10186
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8727280185459015,
+      "learning_rate": 2.898472381320252e-06,
+      "loss": 0.5722,
+      "step": 10187
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7734805303035921,
+      "learning_rate": 2.897536382626609e-06,
+      "loss": 0.4853,
+      "step": 10188
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6444911244666087,
+      "learning_rate": 2.8966004734281216e-06,
+      "loss": 0.4859,
+      "step": 10189
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7303950249005075,
+      "learning_rate": 2.895664653764633e-06,
+      "loss": 0.5195,
+      "step": 10190
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6402140605767778,
+      "learning_rate": 2.8947289236759725e-06,
+      "loss": 0.4515,
+      "step": 10191
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6763932478440389,
+      "learning_rate": 2.8937932832019767e-06,
+      "loss": 0.4783,
+      "step": 10192
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6579200449093628,
+      "learning_rate": 2.8928577323824687e-06,
+      "loss": 0.4732,
+      "step": 10193
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7027925633443014,
+      "learning_rate": 2.8919222712572755e-06,
+      "loss": 0.4976,
+      "step": 10194
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5671219139174667,
+      "learning_rate": 2.8909868998662127e-06,
+      "loss": 0.4658,
+      "step": 10195
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.9705835662620246,
+      "learning_rate": 2.8900516182491002e-06,
+      "loss": 0.5174,
+      "step": 10196
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7697562743226838,
+      "learning_rate": 2.8891164264457457e-06,
+      "loss": 0.4861,
+      "step": 10197
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6050093154754461,
+      "learning_rate": 2.8881813244959612e-06,
+      "loss": 0.5307,
+      "step": 10198
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8398249656435676,
+      "learning_rate": 2.8872463124395457e-06,
+      "loss": 0.5081,
+      "step": 10199
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5772250485459917,
+      "learning_rate": 2.8863113903163078e-06,
+      "loss": 0.4704,
+      "step": 10200
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.679098520169189,
+      "learning_rate": 2.8853765581660365e-06,
+      "loss": 0.5039,
+      "step": 10201
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6360127132472283,
+      "learning_rate": 2.8844418160285305e-06,
+      "loss": 0.4763,
+      "step": 10202
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7243562471193774,
+      "learning_rate": 2.883507163943574e-06,
+      "loss": 0.5064,
+      "step": 10203
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7163903348118532,
+      "learning_rate": 2.882572601950956e-06,
+      "loss": 0.4583,
+      "step": 10204
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5711941198400342,
+      "learning_rate": 2.881638130090454e-06,
+      "loss": 0.4582,
+      "step": 10205
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6274230845694725,
+      "learning_rate": 2.8807037484018496e-06,
+      "loss": 0.4164,
+      "step": 10206
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6260335194664035,
+      "learning_rate": 2.8797694569249122e-06,
+      "loss": 0.5005,
+      "step": 10207
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7128907202421917,
+      "learning_rate": 2.878835255699416e-06,
+      "loss": 0.533,
+      "step": 10208
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7845323777474015,
+      "learning_rate": 2.8779011447651227e-06,
+      "loss": 0.5281,
+      "step": 10209
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.747650927201233,
+      "learning_rate": 2.876967124161798e-06,
+      "loss": 0.5459,
+      "step": 10210
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7193081240913091,
+      "learning_rate": 2.876033193929197e-06,
+      "loss": 0.5552,
+      "step": 10211
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7914127652799089,
+      "learning_rate": 2.8750993541070753e-06,
+      "loss": 0.5435,
+      "step": 10212
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7084305546304169,
+      "learning_rate": 2.874165604735184e-06,
+      "loss": 0.5492,
+      "step": 10213
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5880303783712654,
+      "learning_rate": 2.8732319458532722e-06,
+      "loss": 0.4228,
+      "step": 10214
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6229318713636494,
+      "learning_rate": 2.8722983775010775e-06,
+      "loss": 0.4712,
+      "step": 10215
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6756154003694903,
+      "learning_rate": 2.871364899718344e-06,
+      "loss": 0.503,
+      "step": 10216
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7218600006010616,
+      "learning_rate": 2.8704315125448027e-06,
+      "loss": 0.5003,
+      "step": 10217
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6322725850395896,
+      "learning_rate": 2.869498216020189e-06,
+      "loss": 0.4387,
+      "step": 10218
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6896654022079284,
+      "learning_rate": 2.8685650101842257e-06,
+      "loss": 0.4967,
+      "step": 10219
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7053372394926651,
+      "learning_rate": 2.8676318950766413e-06,
+      "loss": 0.4903,
+      "step": 10220
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6199428687595725,
+      "learning_rate": 2.8666988707371507e-06,
+      "loss": 0.4807,
+      "step": 10221
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6769955706031948,
+      "learning_rate": 2.8657659372054743e-06,
+      "loss": 0.4677,
+      "step": 10222
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.675305212036312,
+      "learning_rate": 2.86483309452132e-06,
+      "loss": 0.5068,
+      "step": 10223
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8707066448333448,
+      "learning_rate": 2.8639003427243974e-06,
+      "loss": 0.4643,
+      "step": 10224
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6364961778623395,
+      "learning_rate": 2.8629676818544135e-06,
+      "loss": 0.4276,
+      "step": 10225
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6801140971692162,
+      "learning_rate": 2.862035111951064e-06,
+      "loss": 0.4505,
+      "step": 10226
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6604427994928885,
+      "learning_rate": 2.86110263305405e-06,
+      "loss": 0.4693,
+      "step": 10227
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7986591803117429,
+      "learning_rate": 2.860170245203061e-06,
+      "loss": 0.4842,
+      "step": 10228
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5920876835762492,
+      "learning_rate": 2.8592379484377874e-06,
+      "loss": 0.4548,
+      "step": 10229
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5540399558151131,
+      "learning_rate": 2.8583057427979123e-06,
+      "loss": 0.4252,
+      "step": 10230
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6078661459183704,
+      "learning_rate": 2.8573736283231172e-06,
+      "loss": 0.4432,
+      "step": 10231
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6682072685001961,
+      "learning_rate": 2.8564416050530826e-06,
+      "loss": 0.4582,
+      "step": 10232
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5689592425546806,
+      "learning_rate": 2.8555096730274768e-06,
+      "loss": 0.4613,
+      "step": 10233
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7428411694563717,
+      "learning_rate": 2.854577832285973e-06,
+      "loss": 0.4917,
+      "step": 10234
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7442687694616843,
+      "learning_rate": 2.8536460828682335e-06,
+      "loss": 0.5116,
+      "step": 10235
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.848214475523199,
+      "learning_rate": 2.852714424813921e-06,
+      "loss": 0.5252,
+      "step": 10236
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6479163829099377,
+      "learning_rate": 2.851782858162696e-06,
+      "loss": 0.5247,
+      "step": 10237
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7847061284805814,
+      "learning_rate": 2.8508513829542086e-06,
+      "loss": 0.5009,
+      "step": 10238
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.832324596194084,
+      "learning_rate": 2.8499199992281113e-06,
+      "loss": 0.5098,
+      "step": 10239
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5739694357939171,
+      "learning_rate": 2.848988707024047e-06,
+      "loss": 0.4637,
+      "step": 10240
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.022437639239965,
+      "learning_rate": 2.8480575063816617e-06,
+      "loss": 0.5215,
+      "step": 10241
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6254561500669832,
+      "learning_rate": 2.8471263973405904e-06,
+      "loss": 0.4251,
+      "step": 10242
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8074913719827148,
+      "learning_rate": 2.8461953799404702e-06,
+      "loss": 0.4465,
+      "step": 10243
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7669159011448926,
+      "learning_rate": 2.8452644542209283e-06,
+      "loss": 0.4949,
+      "step": 10244
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6114192213671742,
+      "learning_rate": 2.8443336202215953e-06,
+      "loss": 0.5023,
+      "step": 10245
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7261979701571393,
+      "learning_rate": 2.843402877982089e-06,
+      "loss": 0.5405,
+      "step": 10246
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7405986778326757,
+      "learning_rate": 2.842472227542033e-06,
+      "loss": 0.5474,
+      "step": 10247
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7283534771526289,
+      "learning_rate": 2.841541668941038e-06,
+      "loss": 0.551,
+      "step": 10248
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.606013040166489,
+      "learning_rate": 2.840611202218716e-06,
+      "loss": 0.4986,
+      "step": 10249
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6612878555801732,
+      "learning_rate": 2.8396808274146757e-06,
+      "loss": 0.5319,
+      "step": 10250
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8857413887674892,
+      "learning_rate": 2.8387505445685205e-06,
+      "loss": 0.5321,
+      "step": 10251
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5987038658197168,
+      "learning_rate": 2.837820353719847e-06,
+      "loss": 0.452,
+      "step": 10252
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6192824010431363,
+      "learning_rate": 2.836890254908252e-06,
+      "loss": 0.5001,
+      "step": 10253
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7636759672701975,
+      "learning_rate": 2.835960248173326e-06,
+      "loss": 0.5391,
+      "step": 10254
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6316826054431608,
+      "learning_rate": 2.835030333554659e-06,
+      "loss": 0.4614,
+      "step": 10255
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6172647901463921,
+      "learning_rate": 2.8341005110918298e-06,
+      "loss": 0.4436,
+      "step": 10256
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6877274755177915,
+      "learning_rate": 2.833170780824423e-06,
+      "loss": 0.498,
+      "step": 10257
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5982828188157019,
+      "learning_rate": 2.83224114279201e-06,
+      "loss": 0.4566,
+      "step": 10258
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.792733103944545,
+      "learning_rate": 2.831311597034166e-06,
+      "loss": 0.4908,
+      "step": 10259
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5374601529386345,
+      "learning_rate": 2.8303821435904554e-06,
+      "loss": 0.4447,
+      "step": 10260
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.627016646477181,
+      "learning_rate": 2.829452782500444e-06,
+      "loss": 0.4607,
+      "step": 10261
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7128397596507715,
+      "learning_rate": 2.828523513803691e-06,
+      "loss": 0.5326,
+      "step": 10262
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6797515649418353,
+      "learning_rate": 2.827594337539755e-06,
+      "loss": 0.4838,
+      "step": 10263
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6103839008157308,
+      "learning_rate": 2.8266652537481842e-06,
+      "loss": 0.4159,
+      "step": 10264
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6107744884089784,
+      "learning_rate": 2.8257362624685316e-06,
+      "loss": 0.4653,
+      "step": 10265
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6060263568691121,
+      "learning_rate": 2.824807363740335e-06,
+      "loss": 0.4345,
+      "step": 10266
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6151861276284115,
+      "learning_rate": 2.82387855760314e-06,
+      "loss": 0.4539,
+      "step": 10267
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.719312302955807,
+      "learning_rate": 2.8229498440964797e-06,
+      "loss": 0.5098,
+      "step": 10268
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7787548917253838,
+      "learning_rate": 2.82202122325989e-06,
+      "loss": 0.4903,
+      "step": 10269
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6801256709394332,
+      "learning_rate": 2.821092695132894e-06,
+      "loss": 0.4544,
+      "step": 10270
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8934962559810521,
+      "learning_rate": 2.8201642597550216e-06,
+      "loss": 0.4587,
+      "step": 10271
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5458415199796832,
+      "learning_rate": 2.8192359171657883e-06,
+      "loss": 0.4295,
+      "step": 10272
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5869432881126402,
+      "learning_rate": 2.818307667404716e-06,
+      "loss": 0.4087,
+      "step": 10273
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8465861078488011,
+      "learning_rate": 2.8173795105113105e-06,
+      "loss": 0.4923,
+      "step": 10274
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6617769884192014,
+      "learning_rate": 2.8164514465250887e-06,
+      "loss": 0.4302,
+      "step": 10275
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7092001774204404,
+      "learning_rate": 2.8155234754855486e-06,
+      "loss": 0.5092,
+      "step": 10276
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7022471518079075,
+      "learning_rate": 2.8145955974321957e-06,
+      "loss": 0.5295,
+      "step": 10277
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8104323622871926,
+      "learning_rate": 2.8136678124045234e-06,
+      "loss": 0.4892,
+      "step": 10278
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6373869737959377,
+      "learning_rate": 2.8127401204420276e-06,
+      "loss": 0.4946,
+      "step": 10279
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7387560399374357,
+      "learning_rate": 2.811812521584193e-06,
+      "loss": 0.4795,
+      "step": 10280
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5850546118255726,
+      "learning_rate": 2.8108850158705093e-06,
+      "loss": 0.5128,
+      "step": 10281
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6728292164626909,
+      "learning_rate": 2.8099576033404524e-06,
+      "loss": 0.4763,
+      "step": 10282
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.684555096851381,
+      "learning_rate": 2.8090302840335043e-06,
+      "loss": 0.4436,
+      "step": 10283
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6571009118436286,
+      "learning_rate": 2.808103057989134e-06,
+      "loss": 0.4849,
+      "step": 10284
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7087608217287694,
+      "learning_rate": 2.8071759252468138e-06,
+      "loss": 0.5212,
+      "step": 10285
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8403310059335515,
+      "learning_rate": 2.8062488858460056e-06,
+      "loss": 0.5864,
+      "step": 10286
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7314393986426435,
+      "learning_rate": 2.8053219398261715e-06,
+      "loss": 0.4602,
+      "step": 10287
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7083839069407044,
+      "learning_rate": 2.8043950872267717e-06,
+      "loss": 0.4476,
+      "step": 10288
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6700185312252959,
+      "learning_rate": 2.8034683280872544e-06,
+      "loss": 0.4606,
+      "step": 10289
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6161550710054887,
+      "learning_rate": 2.8025416624470737e-06,
+      "loss": 0.4463,
+      "step": 10290
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5871924184064675,
+      "learning_rate": 2.8016150903456706e-06,
+      "loss": 0.4657,
+      "step": 10291
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7160656083847337,
+      "learning_rate": 2.8006886118224898e-06,
+      "loss": 0.5152,
+      "step": 10292
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8708033483477031,
+      "learning_rate": 2.7997622269169643e-06,
+      "loss": 0.5324,
+      "step": 10293
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8495864015237135,
+      "learning_rate": 2.7988359356685307e-06,
+      "loss": 0.5304,
+      "step": 10294
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8758556575715707,
+      "learning_rate": 2.7979097381166185e-06,
+      "loss": 0.483,
+      "step": 10295
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6023690437847421,
+      "learning_rate": 2.7969836343006507e-06,
+      "loss": 0.4535,
+      "step": 10296
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7341064892883139,
+      "learning_rate": 2.796057624260051e-06,
+      "loss": 0.5287,
+      "step": 10297
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6943803505952622,
+      "learning_rate": 2.7951317080342345e-06,
+      "loss": 0.4895,
+      "step": 10298
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5712222322452283,
+      "learning_rate": 2.7942058856626147e-06,
+      "loss": 0.3958,
+      "step": 10299
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.680760703570845,
+      "learning_rate": 2.7932801571846035e-06,
+      "loss": 0.5178,
+      "step": 10300
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7785017709815346,
+      "learning_rate": 2.792354522639603e-06,
+      "loss": 0.5255,
+      "step": 10301
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8857959021537309,
+      "learning_rate": 2.791428982067017e-06,
+      "loss": 0.4876,
+      "step": 10302
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7026159846106256,
+      "learning_rate": 2.7905035355062403e-06,
+      "loss": 0.5265,
+      "step": 10303
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6851347398625993,
+      "learning_rate": 2.7895781829966696e-06,
+      "loss": 0.5018,
+      "step": 10304
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5777702955082746,
+      "learning_rate": 2.7886529245776896e-06,
+      "loss": 0.4829,
+      "step": 10305
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6512647504553011,
+      "learning_rate": 2.7877277602886908e-06,
+      "loss": 0.5077,
+      "step": 10306
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8217845487809845,
+      "learning_rate": 2.786802690169049e-06,
+      "loss": 0.5281,
+      "step": 10307
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6647988542616354,
+      "learning_rate": 2.7858777142581474e-06,
+      "loss": 0.5418,
+      "step": 10308
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8704886978408509,
+      "learning_rate": 2.7849528325953532e-06,
+      "loss": 0.5389,
+      "step": 10309
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7667554399995513,
+      "learning_rate": 2.784028045220041e-06,
+      "loss": 0.5095,
+      "step": 10310
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5799480689458069,
+      "learning_rate": 2.78310335217157e-06,
+      "loss": 0.4272,
+      "step": 10311
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6003253809909085,
+      "learning_rate": 2.7821787534893086e-06,
+      "loss": 0.4406,
+      "step": 10312
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6632551367750247,
+      "learning_rate": 2.781254249212609e-06,
+      "loss": 0.4135,
+      "step": 10313
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6204520004420718,
+      "learning_rate": 2.7803298393808275e-06,
+      "loss": 0.5279,
+      "step": 10314
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.718102852184217,
+      "learning_rate": 2.7794055240333097e-06,
+      "loss": 0.5138,
+      "step": 10315
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7621876313546597,
+      "learning_rate": 2.778481303209405e-06,
+      "loss": 0.4995,
+      "step": 10316
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6308741026634868,
+      "learning_rate": 2.77755717694845e-06,
+      "loss": 0.477,
+      "step": 10317
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5793751279011736,
+      "learning_rate": 2.7766331452897866e-06,
+      "loss": 0.4422,
+      "step": 10318
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7060617440529559,
+      "learning_rate": 2.775709208272742e-06,
+      "loss": 0.4876,
+      "step": 10319
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.9337510585974735,
+      "learning_rate": 2.774785365936652e-06,
+      "loss": 0.5147,
+      "step": 10320
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5594443465244961,
+      "learning_rate": 2.7738616183208355e-06,
+      "loss": 0.4208,
+      "step": 10321
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5780470233219226,
+      "learning_rate": 2.772937965464619e-06,
+      "loss": 0.4714,
+      "step": 10322
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5957072811076571,
+      "learning_rate": 2.772014407407312e-06,
+      "loss": 0.447,
+      "step": 10323
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6632208242494634,
+      "learning_rate": 2.771090944188236e-06,
+      "loss": 0.5158,
+      "step": 10324
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7936907643249319,
+      "learning_rate": 2.770167575846694e-06,
+      "loss": 0.5607,
+      "step": 10325
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0528812035680306,
+      "learning_rate": 2.769244302421995e-06,
+      "loss": 0.528,
+      "step": 10326
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5810315354662225,
+      "learning_rate": 2.7683211239534346e-06,
+      "loss": 0.4219,
+      "step": 10327
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6969253167692392,
+      "learning_rate": 2.7673980404803156e-06,
+      "loss": 0.5126,
+      "step": 10328
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7726218100733003,
+      "learning_rate": 2.766475052041926e-06,
+      "loss": 0.5371,
+      "step": 10329
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7099240571205144,
+      "learning_rate": 2.765552158677557e-06,
+      "loss": 0.4905,
+      "step": 10330
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.619004841127661,
+      "learning_rate": 2.7646293604264908e-06,
+      "loss": 0.4465,
+      "step": 10331
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6415019938912052,
+      "learning_rate": 2.763706657328011e-06,
+      "loss": 0.5054,
+      "step": 10332
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6756060090921461,
+      "learning_rate": 2.7627840494213914e-06,
+      "loss": 0.4842,
+      "step": 10333
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6233721176824821,
+      "learning_rate": 2.761861536745908e-06,
+      "loss": 0.5191,
+      "step": 10334
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6540391906862236,
+      "learning_rate": 2.7609391193408243e-06,
+      "loss": 0.4916,
+      "step": 10335
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5785833943689869,
+      "learning_rate": 2.760016797245407e-06,
+      "loss": 0.4602,
+      "step": 10336
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6214001281557133,
+      "learning_rate": 2.7590945704989168e-06,
+      "loss": 0.4682,
+      "step": 10337
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7143726147005993,
+      "learning_rate": 2.758172439140612e-06,
+      "loss": 0.515,
+      "step": 10338
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6172944417762497,
+      "learning_rate": 2.7572504032097406e-06,
+      "loss": 0.4129,
+      "step": 10339
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6894994342783959,
+      "learning_rate": 2.7563284627455545e-06,
+      "loss": 0.4387,
+      "step": 10340
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6526816237412422,
+      "learning_rate": 2.7554066177872948e-06,
+      "loss": 0.4414,
+      "step": 10341
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6287719420341326,
+      "learning_rate": 2.754484868374204e-06,
+      "loss": 0.4557,
+      "step": 10342
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8484841580775698,
+      "learning_rate": 2.753563214545515e-06,
+      "loss": 0.4992,
+      "step": 10343
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6082974126964179,
+      "learning_rate": 2.752641656340463e-06,
+      "loss": 0.4208,
+      "step": 10344
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.594648707224296,
+      "learning_rate": 2.7517201937982724e-06,
+      "loss": 0.4778,
+      "step": 10345
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7926787184939269,
+      "learning_rate": 2.750798826958171e-06,
+      "loss": 0.4674,
+      "step": 10346
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6785821081690722,
+      "learning_rate": 2.749877555859373e-06,
+      "loss": 0.4938,
+      "step": 10347
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.720702003143994,
+      "learning_rate": 2.748956380541098e-06,
+      "loss": 0.5642,
+      "step": 10348
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7127434240465371,
+      "learning_rate": 2.7480353010425586e-06,
+      "loss": 0.5575,
+      "step": 10349
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8407082801634858,
+      "learning_rate": 2.747114317402958e-06,
+      "loss": 0.4769,
+      "step": 10350
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7015039866122491,
+      "learning_rate": 2.7461934296615034e-06,
+      "loss": 0.4897,
+      "step": 10351
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7008876470538534,
+      "learning_rate": 2.7452726378573913e-06,
+      "loss": 0.4816,
+      "step": 10352
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6092498228414746,
+      "learning_rate": 2.7443519420298192e-06,
+      "loss": 0.4398,
+      "step": 10353
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6386698963862778,
+      "learning_rate": 2.743431342217975e-06,
+      "loss": 0.4287,
+      "step": 10354
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6112131601942666,
+      "learning_rate": 2.742510838461048e-06,
+      "loss": 0.4682,
+      "step": 10355
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6570679633884028,
+      "learning_rate": 2.741590430798222e-06,
+      "loss": 0.4796,
+      "step": 10356
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.569073328216951,
+      "learning_rate": 2.7406701192686725e-06,
+      "loss": 0.4368,
+      "step": 10357
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5715520359315882,
+      "learning_rate": 2.7397499039115786e-06,
+      "loss": 0.4126,
+      "step": 10358
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7368551643100076,
+      "learning_rate": 2.7388297847661066e-06,
+      "loss": 0.5256,
+      "step": 10359
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7238767532920479,
+      "learning_rate": 2.737909761871425e-06,
+      "loss": 0.5203,
+      "step": 10360
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5591126185576063,
+      "learning_rate": 2.736989835266698e-06,
+      "loss": 0.4331,
+      "step": 10361
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.879536262743953,
+      "learning_rate": 2.7360700049910806e-06,
+      "loss": 0.5727,
+      "step": 10362
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6834681890321648,
+      "learning_rate": 2.7351502710837303e-06,
+      "loss": 0.4925,
+      "step": 10363
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8807061460864501,
+      "learning_rate": 2.734230633583793e-06,
+      "loss": 0.5137,
+      "step": 10364
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7100398944654575,
+      "learning_rate": 2.7333110925304207e-06,
+      "loss": 0.5044,
+      "step": 10365
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8074920784042978,
+      "learning_rate": 2.732391647962749e-06,
+      "loss": 0.548,
+      "step": 10366
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.772978060469139,
+      "learning_rate": 2.7314722999199206e-06,
+      "loss": 0.51,
+      "step": 10367
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6521215880523424,
+      "learning_rate": 2.7305530484410646e-06,
+      "loss": 0.5046,
+      "step": 10368
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6954414989450015,
+      "learning_rate": 2.7296338935653167e-06,
+      "loss": 0.4769,
+      "step": 10369
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8740772687049388,
+      "learning_rate": 2.728714835331796e-06,
+      "loss": 0.5171,
+      "step": 10370
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7581098288126822,
+      "learning_rate": 2.7277958737796283e-06,
+      "loss": 0.5144,
+      "step": 10371
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8589557568648777,
+      "learning_rate": 2.726877008947928e-06,
+      "loss": 0.5091,
+      "step": 10372
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6110068588269921,
+      "learning_rate": 2.7259582408758096e-06,
+      "loss": 0.4446,
+      "step": 10373
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7147663911311242,
+      "learning_rate": 2.725039569602382e-06,
+      "loss": 0.4835,
+      "step": 10374
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7413469584473779,
+      "learning_rate": 2.724120995166752e-06,
+      "loss": 0.544,
+      "step": 10375
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8872715134898821,
+      "learning_rate": 2.723202517608017e-06,
+      "loss": 0.5739,
+      "step": 10376
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6907243464323122,
+      "learning_rate": 2.7222841369652764e-06,
+      "loss": 0.4808,
+      "step": 10377
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5595479299385279,
+      "learning_rate": 2.7213658532776197e-06,
+      "loss": 0.4381,
+      "step": 10378
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6354036481681429,
+      "learning_rate": 2.7204476665841395e-06,
+      "loss": 0.4569,
+      "step": 10379
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6061294230134897,
+      "learning_rate": 2.719529576923916e-06,
+      "loss": 0.4636,
+      "step": 10380
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6723032050636789,
+      "learning_rate": 2.7186115843360326e-06,
+      "loss": 0.4954,
+      "step": 10381
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6208787356448009,
+      "learning_rate": 2.7176936888595625e-06,
+      "loss": 0.488,
+      "step": 10382
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6094514859095329,
+      "learning_rate": 2.71677589053358e-06,
+      "loss": 0.424,
+      "step": 10383
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7362013402079931,
+      "learning_rate": 2.7158581893971504e-06,
+      "loss": 0.4554,
+      "step": 10384
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7158430907069553,
+      "learning_rate": 2.714940585489338e-06,
+      "loss": 0.541,
+      "step": 10385
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7508972989594184,
+      "learning_rate": 2.714023078849203e-06,
+      "loss": 0.5434,
+      "step": 10386
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6837608999964359,
+      "learning_rate": 2.7131056695158033e-06,
+      "loss": 0.493,
+      "step": 10387
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8794385664082642,
+      "learning_rate": 2.7121883575281844e-06,
+      "loss": 0.5156,
+      "step": 10388
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8270004047901359,
+      "learning_rate": 2.7112711429253987e-06,
+      "loss": 0.5802,
+      "step": 10389
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.852907495806386,
+      "learning_rate": 2.7103540257464845e-06,
+      "loss": 0.519,
+      "step": 10390
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6275563166804927,
+      "learning_rate": 2.709437006030485e-06,
+      "loss": 0.5077,
+      "step": 10391
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7216492105876493,
+      "learning_rate": 2.7085200838164306e-06,
+      "loss": 0.5058,
+      "step": 10392
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7367917113004402,
+      "learning_rate": 2.7076032591433554e-06,
+      "loss": 0.5284,
+      "step": 10393
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.3768037595849074,
+      "learning_rate": 2.706686532050282e-06,
+      "loss": 0.527,
+      "step": 10394
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7261575510917068,
+      "learning_rate": 2.7057699025762372e-06,
+      "loss": 0.5497,
+      "step": 10395
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7461490466559105,
+      "learning_rate": 2.7048533707602343e-06,
+      "loss": 0.5054,
+      "step": 10396
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5935209370164812,
+      "learning_rate": 2.7039369366412906e-06,
+      "loss": 0.3943,
+      "step": 10397
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5945092786602868,
+      "learning_rate": 2.7030206002584102e-06,
+      "loss": 0.4446,
+      "step": 10398
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.3263798578125456,
+      "learning_rate": 2.7021043616506073e-06,
+      "loss": 0.5199,
+      "step": 10399
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6861515727606337,
+      "learning_rate": 2.7011882208568764e-06,
+      "loss": 0.4987,
+      "step": 10400
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7205811719774767,
+      "learning_rate": 2.7002721779162194e-06,
+      "loss": 0.5265,
+      "step": 10401
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7385668990841435,
+      "learning_rate": 2.6993562328676247e-06,
+      "loss": 0.5232,
+      "step": 10402
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.73489183386022,
+      "learning_rate": 2.698440385750085e-06,
+      "loss": 0.4837,
+      "step": 10403
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.595810960905717,
+      "learning_rate": 2.6975246366025822e-06,
+      "loss": 0.4371,
+      "step": 10404
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.60407533029049,
+      "learning_rate": 2.6966089854640987e-06,
+      "loss": 0.4683,
+      "step": 10405
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6359714362799845,
+      "learning_rate": 2.695693432373609e-06,
+      "loss": 0.42,
+      "step": 10406
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6396203805091751,
+      "learning_rate": 2.694777977370088e-06,
+      "loss": 0.4969,
+      "step": 10407
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7495227349500316,
+      "learning_rate": 2.6938626204925e-06,
+      "loss": 0.5232,
+      "step": 10408
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6957898811146143,
+      "learning_rate": 2.692947361779813e-06,
+      "loss": 0.4519,
+      "step": 10409
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6135722733042351,
+      "learning_rate": 2.6920322012709832e-06,
+      "loss": 0.5092,
+      "step": 10410
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8106406537348229,
+      "learning_rate": 2.691117139004966e-06,
+      "loss": 0.5211,
+      "step": 10411
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7911049720073707,
+      "learning_rate": 2.6902021750207176e-06,
+      "loss": 0.5253,
+      "step": 10412
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.590970404960969,
+      "learning_rate": 2.689287309357179e-06,
+      "loss": 0.4332,
+      "step": 10413
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6897002819401162,
+      "learning_rate": 2.688372542053297e-06,
+      "loss": 0.4909,
+      "step": 10414
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6949444258759422,
+      "learning_rate": 2.6874578731480072e-06,
+      "loss": 0.4939,
+      "step": 10415
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5653657736103492,
+      "learning_rate": 2.6865433026802485e-06,
+      "loss": 0.4331,
+      "step": 10416
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6137579517920647,
+      "learning_rate": 2.6856288306889465e-06,
+      "loss": 0.4235,
+      "step": 10417
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5642045045712887,
+      "learning_rate": 2.684714457213029e-06,
+      "loss": 0.4155,
+      "step": 10418
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6865290335860689,
+      "learning_rate": 2.6838001822914205e-06,
+      "loss": 0.4674,
+      "step": 10419
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6881986360685356,
+      "learning_rate": 2.6828860059630355e-06,
+      "loss": 0.4622,
+      "step": 10420
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6616896634586286,
+      "learning_rate": 2.68197192826679e-06,
+      "loss": 0.4406,
+      "step": 10421
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6154911590713704,
+      "learning_rate": 2.6810579492415906e-06,
+      "loss": 0.4032,
+      "step": 10422
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6174447590466567,
+      "learning_rate": 2.6801440689263438e-06,
+      "loss": 0.4651,
+      "step": 10423
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6742644993561953,
+      "learning_rate": 2.679230287359953e-06,
+      "loss": 0.4749,
+      "step": 10424
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6392911119217446,
+      "learning_rate": 2.67831660458131e-06,
+      "loss": 0.4367,
+      "step": 10425
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6761821235044461,
+      "learning_rate": 2.6774030206293132e-06,
+      "loss": 0.4819,
+      "step": 10426
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5601652227907855,
+      "learning_rate": 2.6764895355428456e-06,
+      "loss": 0.4231,
+      "step": 10427
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.1323072996383916,
+      "learning_rate": 2.675576149360795e-06,
+      "loss": 0.498,
+      "step": 10428
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6688631409101287,
+      "learning_rate": 2.674662862122038e-06,
+      "loss": 0.4573,
+      "step": 10429
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5925262351078397,
+      "learning_rate": 2.673749673865455e-06,
+      "loss": 0.4537,
+      "step": 10430
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6078301519512345,
+      "learning_rate": 2.672836584629912e-06,
+      "loss": 0.4297,
+      "step": 10431
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6043560196132154,
+      "learning_rate": 2.6719235944542816e-06,
+      "loss": 0.4425,
+      "step": 10432
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6145893193107074,
+      "learning_rate": 2.6710107033774225e-06,
+      "loss": 0.4801,
+      "step": 10433
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7046347501259793,
+      "learning_rate": 2.6700979114381966e-06,
+      "loss": 0.5022,
+      "step": 10434
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.955439941567224,
+      "learning_rate": 2.6691852186754535e-06,
+      "loss": 0.5311,
+      "step": 10435
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7776049542735206,
+      "learning_rate": 2.6682726251280515e-06,
+      "loss": 0.5564,
+      "step": 10436
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6946366134677137,
+      "learning_rate": 2.6673601308348307e-06,
+      "loss": 0.5111,
+      "step": 10437
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6056262640099894,
+      "learning_rate": 2.6664477358346375e-06,
+      "loss": 0.4693,
+      "step": 10438
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.702928538440643,
+      "learning_rate": 2.6655354401663047e-06,
+      "loss": 0.464,
+      "step": 10439
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7428536543476963,
+      "learning_rate": 2.6646232438686704e-06,
+      "loss": 0.5165,
+      "step": 10440
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7889657852679707,
+      "learning_rate": 2.66371114698056e-06,
+      "loss": 0.5534,
+      "step": 10441
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6419378921457135,
+      "learning_rate": 2.6627991495408025e-06,
+      "loss": 0.4272,
+      "step": 10442
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5456671136530137,
+      "learning_rate": 2.6618872515882143e-06,
+      "loss": 0.4286,
+      "step": 10443
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6160738899883795,
+      "learning_rate": 2.6609754531616162e-06,
+      "loss": 0.4285,
+      "step": 10444
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6922031132322712,
+      "learning_rate": 2.6600637542998164e-06,
+      "loss": 0.4647,
+      "step": 10445
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6074556864594911,
+      "learning_rate": 2.6591521550416275e-06,
+      "loss": 0.4475,
+      "step": 10446
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5960269225081608,
+      "learning_rate": 2.658240655425847e-06,
+      "loss": 0.4518,
+      "step": 10447
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6564091672395928,
+      "learning_rate": 2.6573292554912832e-06,
+      "loss": 0.486,
+      "step": 10448
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7316632297416248,
+      "learning_rate": 2.6564179552767245e-06,
+      "loss": 0.4529,
+      "step": 10449
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6384750104167783,
+      "learning_rate": 2.6555067548209655e-06,
+      "loss": 0.4736,
+      "step": 10450
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9465589381693489,
+      "learning_rate": 2.6545956541627913e-06,
+      "loss": 0.5254,
+      "step": 10451
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7444444643752165,
+      "learning_rate": 2.6536846533409865e-06,
+      "loss": 0.4599,
+      "step": 10452
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.540997955708366,
+      "learning_rate": 2.6527737523943264e-06,
+      "loss": 0.3976,
+      "step": 10453
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6451066744950943,
+      "learning_rate": 2.6518629513615894e-06,
+      "loss": 0.4628,
+      "step": 10454
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7762395218882295,
+      "learning_rate": 2.650952250281541e-06,
+      "loss": 0.4863,
+      "step": 10455
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6056665589043336,
+      "learning_rate": 2.6500416491929505e-06,
+      "loss": 0.4591,
+      "step": 10456
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9170622477340048,
+      "learning_rate": 2.6491311481345763e-06,
+      "loss": 0.5517,
+      "step": 10457
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5882766029963012,
+      "learning_rate": 2.6482207471451775e-06,
+      "loss": 0.4447,
+      "step": 10458
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.790139805967078,
+      "learning_rate": 2.6473104462635048e-06,
+      "loss": 0.468,
+      "step": 10459
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5435428189331143,
+      "learning_rate": 2.646400245528309e-06,
+      "loss": 0.4369,
+      "step": 10460
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6846943788527304,
+      "learning_rate": 2.6454901449783334e-06,
+      "loss": 0.4924,
+      "step": 10461
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9569522481544581,
+      "learning_rate": 2.64458014465232e-06,
+      "loss": 0.545,
+      "step": 10462
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.804079115785682,
+      "learning_rate": 2.6436702445890007e-06,
+      "loss": 0.5081,
+      "step": 10463
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5884263214599849,
+      "learning_rate": 2.6427604448271116e-06,
+      "loss": 0.4055,
+      "step": 10464
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5927593392228628,
+      "learning_rate": 2.6418507454053764e-06,
+      "loss": 0.4662,
+      "step": 10465
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7954120688325812,
+      "learning_rate": 2.6409411463625207e-06,
+      "loss": 0.4887,
+      "step": 10466
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7019531549203019,
+      "learning_rate": 2.64003164773726e-06,
+      "loss": 0.4677,
+      "step": 10467
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7026126361040785,
+      "learning_rate": 2.6391222495683134e-06,
+      "loss": 0.4784,
+      "step": 10468
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8387632757186848,
+      "learning_rate": 2.6382129518943856e-06,
+      "loss": 0.4801,
+      "step": 10469
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8084044790296943,
+      "learning_rate": 2.6373037547541867e-06,
+      "loss": 0.5229,
+      "step": 10470
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6055974131143065,
+      "learning_rate": 2.6363946581864156e-06,
+      "loss": 0.4626,
+      "step": 10471
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7762961284619238,
+      "learning_rate": 2.6354856622297707e-06,
+      "loss": 0.5326,
+      "step": 10472
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6834070519293799,
+      "learning_rate": 2.6345767669229482e-06,
+      "loss": 0.4981,
+      "step": 10473
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.653968676520899,
+      "learning_rate": 2.633667972304631e-06,
+      "loss": 0.435,
+      "step": 10474
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6717962390423486,
+      "learning_rate": 2.6327592784135082e-06,
+      "loss": 0.4299,
+      "step": 10475
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.699017160924299,
+      "learning_rate": 2.631850685288257e-06,
+      "loss": 0.4893,
+      "step": 10476
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6701807020173425,
+      "learning_rate": 2.6309421929675562e-06,
+      "loss": 0.4929,
+      "step": 10477
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6229388228504292,
+      "learning_rate": 2.630033801490074e-06,
+      "loss": 0.4395,
+      "step": 10478
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5556153721486646,
+      "learning_rate": 2.6291255108944794e-06,
+      "loss": 0.4388,
+      "step": 10479
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8806362729927469,
+      "learning_rate": 2.628217321219438e-06,
+      "loss": 0.5149,
+      "step": 10480
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.748900621539695,
+      "learning_rate": 2.6273092325036037e-06,
+      "loss": 0.5093,
+      "step": 10481
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7987917057961558,
+      "learning_rate": 2.6264012447856356e-06,
+      "loss": 0.5104,
+      "step": 10482
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5360859916938244,
+      "learning_rate": 2.625493358104179e-06,
+      "loss": 0.4201,
+      "step": 10483
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.698113049700735,
+      "learning_rate": 2.624585572497883e-06,
+      "loss": 0.4428,
+      "step": 10484
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6595825637544322,
+      "learning_rate": 2.6236778880053903e-06,
+      "loss": 0.4381,
+      "step": 10485
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6149852131666086,
+      "learning_rate": 2.622770304665334e-06,
+      "loss": 0.5021,
+      "step": 10486
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6687978346064894,
+      "learning_rate": 2.621862822516351e-06,
+      "loss": 0.4944,
+      "step": 10487
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5596785788185218,
+      "learning_rate": 2.6209554415970668e-06,
+      "loss": 0.4276,
+      "step": 10488
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6159920787070838,
+      "learning_rate": 2.6200481619461087e-06,
+      "loss": 0.5368,
+      "step": 10489
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.803250683156127,
+      "learning_rate": 2.619140983602093e-06,
+      "loss": 0.5126,
+      "step": 10490
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7072484586432487,
+      "learning_rate": 2.61823390660364e-06,
+      "loss": 0.5449,
+      "step": 10491
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7122089858160523,
+      "learning_rate": 2.617326930989357e-06,
+      "loss": 0.5384,
+      "step": 10492
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.764709926973671,
+      "learning_rate": 2.6164200567978538e-06,
+      "loss": 0.5251,
+      "step": 10493
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7965000517741182,
+      "learning_rate": 2.61551328406773e-06,
+      "loss": 0.5426,
+      "step": 10494
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5722800226789765,
+      "learning_rate": 2.614606612837588e-06,
+      "loss": 0.4667,
+      "step": 10495
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6066833520036666,
+      "learning_rate": 2.613700043146018e-06,
+      "loss": 0.5074,
+      "step": 10496
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.540690502361817,
+      "learning_rate": 2.612793575031611e-06,
+      "loss": 0.4146,
+      "step": 10497
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5691137692350555,
+      "learning_rate": 2.6118872085329545e-06,
+      "loss": 0.4446,
+      "step": 10498
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.660356675828652,
+      "learning_rate": 2.610980943688629e-06,
+      "loss": 0.4777,
+      "step": 10499
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8759596562439051,
+      "learning_rate": 2.6100747805372097e-06,
+      "loss": 0.5395,
+      "step": 10500
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5968988089047996,
+      "learning_rate": 2.609168719117271e-06,
+      "loss": 0.526,
+      "step": 10501
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.9312721545499884,
+      "learning_rate": 2.6082627594673777e-06,
+      "loss": 0.542,
+      "step": 10502
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5907621184336649,
+      "learning_rate": 2.607356901626098e-06,
+      "loss": 0.4629,
+      "step": 10503
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7301453061227231,
+      "learning_rate": 2.6064511456319874e-06,
+      "loss": 0.4549,
+      "step": 10504
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6367422057780036,
+      "learning_rate": 2.605545491523605e-06,
+      "loss": 0.4091,
+      "step": 10505
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6060024480535015,
+      "learning_rate": 2.6046399393394973e-06,
+      "loss": 0.4459,
+      "step": 10506
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.603006322669303,
+      "learning_rate": 2.6037344891182147e-06,
+      "loss": 0.4441,
+      "step": 10507
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7097251674959564,
+      "learning_rate": 2.602829140898296e-06,
+      "loss": 0.4737,
+      "step": 10508
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6045866969731865,
+      "learning_rate": 2.6019238947182802e-06,
+      "loss": 0.5126,
+      "step": 10509
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6806143131924589,
+      "learning_rate": 2.6010187506167008e-06,
+      "loss": 0.5375,
+      "step": 10510
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.735141985519335,
+      "learning_rate": 2.6001137086320895e-06,
+      "loss": 0.5032,
+      "step": 10511
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6999838503513939,
+      "learning_rate": 2.5992087688029666e-06,
+      "loss": 0.5049,
+      "step": 10512
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6874803323596783,
+      "learning_rate": 2.5983039311678567e-06,
+      "loss": 0.4751,
+      "step": 10513
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5926923896376749,
+      "learning_rate": 2.597399195765271e-06,
+      "loss": 0.4414,
+      "step": 10514
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5425670683386655,
+      "learning_rate": 2.5964945626337268e-06,
+      "loss": 0.4205,
+      "step": 10515
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6970937596512157,
+      "learning_rate": 2.5955900318117266e-06,
+      "loss": 0.4235,
+      "step": 10516
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6205585218731938,
+      "learning_rate": 2.5946856033377767e-06,
+      "loss": 0.4693,
+      "step": 10517
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6938979631909559,
+      "learning_rate": 2.593781277250372e-06,
+      "loss": 0.4685,
+      "step": 10518
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7968496125303536,
+      "learning_rate": 2.5928770535880122e-06,
+      "loss": 0.5244,
+      "step": 10519
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7664858655971052,
+      "learning_rate": 2.591972932389182e-06,
+      "loss": 0.5162,
+      "step": 10520
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6411154348649356,
+      "learning_rate": 2.591068913692371e-06,
+      "loss": 0.4921,
+      "step": 10521
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8168666097855334,
+      "learning_rate": 2.5901649975360543e-06,
+      "loss": 0.5143,
+      "step": 10522
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7723265278626535,
+      "learning_rate": 2.5892611839587175e-06,
+      "loss": 0.5433,
+      "step": 10523
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8534329527893085,
+      "learning_rate": 2.588357472998826e-06,
+      "loss": 0.4499,
+      "step": 10524
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8478258952328214,
+      "learning_rate": 2.587453864694852e-06,
+      "loss": 0.5397,
+      "step": 10525
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6214870427952506,
+      "learning_rate": 2.586550359085256e-06,
+      "loss": 0.4357,
+      "step": 10526
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6422511873552887,
+      "learning_rate": 2.5856469562085013e-06,
+      "loss": 0.4906,
+      "step": 10527
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7478310730735366,
+      "learning_rate": 2.584743656103038e-06,
+      "loss": 0.4967,
+      "step": 10528
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6021550539295655,
+      "learning_rate": 2.5838404588073212e-06,
+      "loss": 0.4701,
+      "step": 10529
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5786767914169088,
+      "learning_rate": 2.5829373643597932e-06,
+      "loss": 0.4634,
+      "step": 10530
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5903349405229537,
+      "learning_rate": 2.5820343727989e-06,
+      "loss": 0.4783,
+      "step": 10531
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7419883312138332,
+      "learning_rate": 2.5811314841630746e-06,
+      "loss": 0.5394,
+      "step": 10532
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.642325443778341,
+      "learning_rate": 2.5802286984907544e-06,
+      "loss": 0.4568,
+      "step": 10533
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6013737178069675,
+      "learning_rate": 2.5793260158203637e-06,
+      "loss": 0.4121,
+      "step": 10534
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6248434686109372,
+      "learning_rate": 2.57842343619033e-06,
+      "loss": 0.47,
+      "step": 10535
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6827723721043608,
+      "learning_rate": 2.577520959639074e-06,
+      "loss": 0.5715,
+      "step": 10536
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7917541185057193,
+      "learning_rate": 2.576618586205007e-06,
+      "loss": 0.5129,
+      "step": 10537
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6355583024877731,
+      "learning_rate": 2.5757163159265454e-06,
+      "loss": 0.4837,
+      "step": 10538
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6567955558681297,
+      "learning_rate": 2.574814148842091e-06,
+      "loss": 0.4935,
+      "step": 10539
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8557123190373086,
+      "learning_rate": 2.5739120849900488e-06,
+      "loss": 0.4954,
+      "step": 10540
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7187077331673724,
+      "learning_rate": 2.573010124408818e-06,
+      "loss": 0.4767,
+      "step": 10541
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7783758985801082,
+      "learning_rate": 2.572108267136789e-06,
+      "loss": 0.5312,
+      "step": 10542
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7522295070272268,
+      "learning_rate": 2.5712065132123544e-06,
+      "loss": 0.5364,
+      "step": 10543
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6587259348762554,
+      "learning_rate": 2.570304862673896e-06,
+      "loss": 0.4672,
+      "step": 10544
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8102807501813335,
+      "learning_rate": 2.5694033155597962e-06,
+      "loss": 0.4768,
+      "step": 10545
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7204510924799229,
+      "learning_rate": 2.5685018719084286e-06,
+      "loss": 0.5349,
+      "step": 10546
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6724871736736747,
+      "learning_rate": 2.567600531758167e-06,
+      "loss": 0.4485,
+      "step": 10547
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8287670774381877,
+      "learning_rate": 2.566699295147379e-06,
+      "loss": 0.5701,
+      "step": 10548
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7504033198606173,
+      "learning_rate": 2.5657981621144252e-06,
+      "loss": 0.5133,
+      "step": 10549
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6510095384940816,
+      "learning_rate": 2.5648971326976667e-06,
+      "loss": 0.4398,
+      "step": 10550
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7776999421851082,
+      "learning_rate": 2.5639962069354534e-06,
+      "loss": 0.5608,
+      "step": 10551
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8928014711124543,
+      "learning_rate": 2.56309538486614e-06,
+      "loss": 0.5259,
+      "step": 10552
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6441973623516792,
+      "learning_rate": 2.5621946665280663e-06,
+      "loss": 0.5076,
+      "step": 10553
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7301758034332348,
+      "learning_rate": 2.5612940519595774e-06,
+      "loss": 0.4537,
+      "step": 10554
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6560502718525917,
+      "learning_rate": 2.5603935411990056e-06,
+      "loss": 0.4863,
+      "step": 10555
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.0806354582198998,
+      "learning_rate": 2.5594931342846874e-06,
+      "loss": 0.5566,
+      "step": 10556
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.9032036502313324,
+      "learning_rate": 2.558592831254946e-06,
+      "loss": 0.5064,
+      "step": 10557
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6583914367167848,
+      "learning_rate": 2.557692632148107e-06,
+      "loss": 0.4844,
+      "step": 10558
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.55842124426622,
+      "learning_rate": 2.556792537002485e-06,
+      "loss": 0.4443,
+      "step": 10559
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7925442216745742,
+      "learning_rate": 2.555892545856401e-06,
+      "loss": 0.4937,
+      "step": 10560
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6272378172920655,
+      "learning_rate": 2.5549926587481596e-06,
+      "loss": 0.4897,
+      "step": 10561
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5655246931149864,
+      "learning_rate": 2.554092875716069e-06,
+      "loss": 0.4677,
+      "step": 10562
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.683200082924845,
+      "learning_rate": 2.553193196798427e-06,
+      "loss": 0.5002,
+      "step": 10563
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6882389962346854,
+      "learning_rate": 2.5522936220335337e-06,
+      "loss": 0.4934,
+      "step": 10564
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7629533289317983,
+      "learning_rate": 2.5513941514596766e-06,
+      "loss": 0.5429,
+      "step": 10565
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7484197632077519,
+      "learning_rate": 2.5504947851151486e-06,
+      "loss": 0.5382,
+      "step": 10566
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7874615109126454,
+      "learning_rate": 2.5495955230382275e-06,
+      "loss": 0.5133,
+      "step": 10567
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6060312838473805,
+      "learning_rate": 2.548696365267197e-06,
+      "loss": 0.4946,
+      "step": 10568
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6841542014755738,
+      "learning_rate": 2.547797311840327e-06,
+      "loss": 0.5041,
+      "step": 10569
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.635146392666409,
+      "learning_rate": 2.546898362795891e-06,
+      "loss": 0.4888,
+      "step": 10570
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5969763884589779,
+      "learning_rate": 2.5459995181721493e-06,
+      "loss": 0.4823,
+      "step": 10571
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6279444187968927,
+      "learning_rate": 2.54510077800737e-06,
+      "loss": 0.4376,
+      "step": 10572
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7474066497095003,
+      "learning_rate": 2.5442021423398034e-06,
+      "loss": 0.4676,
+      "step": 10573
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8083824948346612,
+      "learning_rate": 2.5433036112077064e-06,
+      "loss": 0.4662,
+      "step": 10574
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6992772934404534,
+      "learning_rate": 2.5424051846493225e-06,
+      "loss": 0.4861,
+      "step": 10575
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8115358118946803,
+      "learning_rate": 2.541506862702898e-06,
+      "loss": 0.5079,
+      "step": 10576
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5898076156590896,
+      "learning_rate": 2.540608645406668e-06,
+      "loss": 0.4876,
+      "step": 10577
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6591181353526833,
+      "learning_rate": 2.539710532798871e-06,
+      "loss": 0.432,
+      "step": 10578
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6483357111405805,
+      "learning_rate": 2.538812524917733e-06,
+      "loss": 0.4906,
+      "step": 10579
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7201600310027874,
+      "learning_rate": 2.5379146218014828e-06,
+      "loss": 0.5091,
+      "step": 10580
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7601836067444325,
+      "learning_rate": 2.5370168234883375e-06,
+      "loss": 0.5357,
+      "step": 10581
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6314016482768607,
+      "learning_rate": 2.536119130016518e-06,
+      "loss": 0.51,
+      "step": 10582
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6448711450255533,
+      "learning_rate": 2.535221541424231e-06,
+      "loss": 0.5138,
+      "step": 10583
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7122061061104349,
+      "learning_rate": 2.5343240577496864e-06,
+      "loss": 0.5563,
+      "step": 10584
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8436835601495344,
+      "learning_rate": 2.533426679031088e-06,
+      "loss": 0.5324,
+      "step": 10585
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5453104602450376,
+      "learning_rate": 2.532529405306636e-06,
+      "loss": 0.4452,
+      "step": 10586
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6346023638934389,
+      "learning_rate": 2.5316322366145195e-06,
+      "loss": 0.4616,
+      "step": 10587
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7173474582293952,
+      "learning_rate": 2.5307351729929334e-06,
+      "loss": 0.4771,
+      "step": 10588
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5371831843226925,
+      "learning_rate": 2.5298382144800583e-06,
+      "loss": 0.4291,
+      "step": 10589
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7471570013389153,
+      "learning_rate": 2.528941361114079e-06,
+      "loss": 0.5323,
+      "step": 10590
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7050985320255504,
+      "learning_rate": 2.5280446129331675e-06,
+      "loss": 0.4929,
+      "step": 10591
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6876108715515993,
+      "learning_rate": 2.5271479699754996e-06,
+      "loss": 0.487,
+      "step": 10592
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8497204672920528,
+      "learning_rate": 2.5262514322792387e-06,
+      "loss": 0.5103,
+      "step": 10593
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7966053083109497,
+      "learning_rate": 2.525354999882551e-06,
+      "loss": 0.5476,
+      "step": 10594
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6943922207606278,
+      "learning_rate": 2.524458672823592e-06,
+      "loss": 0.5145,
+      "step": 10595
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7621090276845341,
+      "learning_rate": 2.5235624511405153e-06,
+      "loss": 0.5201,
+      "step": 10596
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6480220236340348,
+      "learning_rate": 2.522666334871474e-06,
+      "loss": 0.4302,
+      "step": 10597
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5352004794161997,
+      "learning_rate": 2.5217703240546087e-06,
+      "loss": 0.4377,
+      "step": 10598
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6158427016130312,
+      "learning_rate": 2.520874418728063e-06,
+      "loss": 0.4572,
+      "step": 10599
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6290232580546541,
+      "learning_rate": 2.5199786189299703e-06,
+      "loss": 0.4678,
+      "step": 10600
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8433508395950833,
+      "learning_rate": 2.519082924698464e-06,
+      "loss": 0.5943,
+      "step": 10601
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7075211559000786,
+      "learning_rate": 2.5181873360716667e-06,
+      "loss": 0.5505,
+      "step": 10602
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7983628484356027,
+      "learning_rate": 2.5172918530877048e-06,
+      "loss": 0.5217,
+      "step": 10603
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6361455888035236,
+      "learning_rate": 2.516396475784697e-06,
+      "loss": 0.4441,
+      "step": 10604
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6358294421365027,
+      "learning_rate": 2.515501204200753e-06,
+      "loss": 0.4485,
+      "step": 10605
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6299572756032057,
+      "learning_rate": 2.514606038373985e-06,
+      "loss": 0.4884,
+      "step": 10606
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7507582170250209,
+      "learning_rate": 2.513710978342494e-06,
+      "loss": 0.4692,
+      "step": 10607
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6674577672579849,
+      "learning_rate": 2.512816024144382e-06,
+      "loss": 0.4979,
+      "step": 10608
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7385789629712277,
+      "learning_rate": 2.5119211758177454e-06,
+      "loss": 0.5023,
+      "step": 10609
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7671958718927128,
+      "learning_rate": 2.5110264334006714e-06,
+      "loss": 0.524,
+      "step": 10610
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.1830286879294056,
+      "learning_rate": 2.5101317969312505e-06,
+      "loss": 0.4849,
+      "step": 10611
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6926632994338192,
+      "learning_rate": 2.5092372664475607e-06,
+      "loss": 0.4315,
+      "step": 10612
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6506380234794855,
+      "learning_rate": 2.5083428419876833e-06,
+      "loss": 0.4335,
+      "step": 10613
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6677770486213096,
+      "learning_rate": 2.5074485235896873e-06,
+      "loss": 0.4615,
+      "step": 10614
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6422924304526997,
+      "learning_rate": 2.5065543112916434e-06,
+      "loss": 0.396,
+      "step": 10615
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6540009378319542,
+      "learning_rate": 2.5056602051316136e-06,
+      "loss": 0.4973,
+      "step": 10616
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7575700931995678,
+      "learning_rate": 2.5047662051476597e-06,
+      "loss": 0.5024,
+      "step": 10617
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6582059655711523,
+      "learning_rate": 2.503872311377833e-06,
+      "loss": 0.4495,
+      "step": 10618
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6696795396076224,
+      "learning_rate": 2.502978523860187e-06,
+      "loss": 0.4741,
+      "step": 10619
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6435913695776708,
+      "learning_rate": 2.502084842632765e-06,
+      "loss": 0.483,
+      "step": 10620
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.730353290183102,
+      "learning_rate": 2.501191267733608e-06,
+      "loss": 0.4844,
+      "step": 10621
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6335069027394884,
+      "learning_rate": 2.5002977992007538e-06,
+      "loss": 0.4521,
+      "step": 10622
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6154280973499405,
+      "learning_rate": 2.499404437072237e-06,
+      "loss": 0.4934,
+      "step": 10623
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5850241018083346,
+      "learning_rate": 2.4985111813860803e-06,
+      "loss": 0.4782,
+      "step": 10624
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6933493158022256,
+      "learning_rate": 2.4976180321803107e-06,
+      "loss": 0.4494,
+      "step": 10625
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6635587136374863,
+      "learning_rate": 2.4967249894929436e-06,
+      "loss": 0.4941,
+      "step": 10626
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.625311579433706,
+      "learning_rate": 2.4958320533619956e-06,
+      "loss": 0.4488,
+      "step": 10627
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6327476384586453,
+      "learning_rate": 2.494939223825473e-06,
+      "loss": 0.4513,
+      "step": 10628
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.773105806783291,
+      "learning_rate": 2.494046500921385e-06,
+      "loss": 0.5474,
+      "step": 10629
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6091867867165777,
+      "learning_rate": 2.4931538846877278e-06,
+      "loss": 0.4876,
+      "step": 10630
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7491279739456798,
+      "learning_rate": 2.4922613751625003e-06,
+      "loss": 0.4838,
+      "step": 10631
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6898837828551352,
+      "learning_rate": 2.4913689723836908e-06,
+      "loss": 0.4972,
+      "step": 10632
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6778226639889613,
+      "learning_rate": 2.490476676389288e-06,
+      "loss": 0.5227,
+      "step": 10633
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7018927983205488,
+      "learning_rate": 2.4895844872172737e-06,
+      "loss": 0.5119,
+      "step": 10634
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8155583147329938,
+      "learning_rate": 2.4886924049056273e-06,
+      "loss": 0.5657,
+      "step": 10635
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8061910783214565,
+      "learning_rate": 2.4878004294923187e-06,
+      "loss": 0.4525,
+      "step": 10636
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5574081462277762,
+      "learning_rate": 2.4869085610153193e-06,
+      "loss": 0.4306,
+      "step": 10637
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6670760635317258,
+      "learning_rate": 2.48601679951259e-06,
+      "loss": 0.4747,
+      "step": 10638
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.570575085827016,
+      "learning_rate": 2.485125145022094e-06,
+      "loss": 0.4678,
+      "step": 10639
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7311714371413807,
+      "learning_rate": 2.4842335975817825e-06,
+      "loss": 0.5276,
+      "step": 10640
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7817605699699901,
+      "learning_rate": 2.4833421572296094e-06,
+      "loss": 0.5376,
+      "step": 10641
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7084791512606515,
+      "learning_rate": 2.4824508240035167e-06,
+      "loss": 0.459,
+      "step": 10642
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5426925526301956,
+      "learning_rate": 2.481559597941449e-06,
+      "loss": 0.4346,
+      "step": 10643
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7486718158103108,
+      "learning_rate": 2.48066847908134e-06,
+      "loss": 0.4588,
+      "step": 10644
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7632464265884233,
+      "learning_rate": 2.479777467461125e-06,
+      "loss": 0.4987,
+      "step": 10645
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6019727398459266,
+      "learning_rate": 2.478886563118726e-06,
+      "loss": 0.4955,
+      "step": 10646
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5631170630230982,
+      "learning_rate": 2.477995766092073e-06,
+      "loss": 0.4184,
+      "step": 10647
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6501602303015634,
+      "learning_rate": 2.4771050764190795e-06,
+      "loss": 0.4438,
+      "step": 10648
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6297624260295119,
+      "learning_rate": 2.4762144941376627e-06,
+      "loss": 0.457,
+      "step": 10649
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.604528205843405,
+      "learning_rate": 2.475324019285728e-06,
+      "loss": 0.4524,
+      "step": 10650
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5504661044380628,
+      "learning_rate": 2.4744336519011848e-06,
+      "loss": 0.4308,
+      "step": 10651
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7918434571931015,
+      "learning_rate": 2.4735433920219275e-06,
+      "loss": 0.4656,
+      "step": 10652
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5493917510346389,
+      "learning_rate": 2.472653239685857e-06,
+      "loss": 0.3887,
+      "step": 10653
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5945150597257793,
+      "learning_rate": 2.4717631949308603e-06,
+      "loss": 0.4757,
+      "step": 10654
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7060958545896473,
+      "learning_rate": 2.4708732577948263e-06,
+      "loss": 0.5,
+      "step": 10655
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.530397584421947,
+      "learning_rate": 2.4699834283156342e-06,
+      "loss": 0.4229,
+      "step": 10656
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7672624101885411,
+      "learning_rate": 2.4690937065311647e-06,
+      "loss": 0.506,
+      "step": 10657
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7874569311621559,
+      "learning_rate": 2.468204092479287e-06,
+      "loss": 0.5733,
+      "step": 10658
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8460276648134271,
+      "learning_rate": 2.4673145861978705e-06,
+      "loss": 0.5846,
+      "step": 10659
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7828687831319213,
+      "learning_rate": 2.466425187724781e-06,
+      "loss": 0.5239,
+      "step": 10660
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7700224106343307,
+      "learning_rate": 2.465535897097872e-06,
+      "loss": 0.4734,
+      "step": 10661
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5346197722840267,
+      "learning_rate": 2.464646714355004e-06,
+      "loss": 0.4442,
+      "step": 10662
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6108654957329758,
+      "learning_rate": 2.4637576395340214e-06,
+      "loss": 0.4543,
+      "step": 10663
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7150591385887055,
+      "learning_rate": 2.4628686726727708e-06,
+      "loss": 0.5846,
+      "step": 10664
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7584122496864705,
+      "learning_rate": 2.4619798138090954e-06,
+      "loss": 0.5119,
+      "step": 10665
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7356999300457018,
+      "learning_rate": 2.4610910629808273e-06,
+      "loss": 0.4591,
+      "step": 10666
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6135349839298728,
+      "learning_rate": 2.4602024202258013e-06,
+      "loss": 0.5139,
+      "step": 10667
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.801710744306865,
+      "learning_rate": 2.4593138855818407e-06,
+      "loss": 0.5061,
+      "step": 10668
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6001626031433973,
+      "learning_rate": 2.4584254590867702e-06,
+      "loss": 0.5057,
+      "step": 10669
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7235302494936137,
+      "learning_rate": 2.457537140778405e-06,
+      "loss": 0.5199,
+      "step": 10670
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.603529320577914,
+      "learning_rate": 2.456648930694559e-06,
+      "loss": 0.4864,
+      "step": 10671
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6812204286898121,
+      "learning_rate": 2.4557608288730424e-06,
+      "loss": 0.5462,
+      "step": 10672
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7765607089730986,
+      "learning_rate": 2.454872835351655e-06,
+      "loss": 0.4848,
+      "step": 10673
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7113788664381167,
+      "learning_rate": 2.4539849501682e-06,
+      "loss": 0.4204,
+      "step": 10674
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6469614434228078,
+      "learning_rate": 2.453097173360468e-06,
+      "loss": 0.452,
+      "step": 10675
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6888606628738865,
+      "learning_rate": 2.4522095049662526e-06,
+      "loss": 0.4846,
+      "step": 10676
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7524738566659702,
+      "learning_rate": 2.451321945023335e-06,
+      "loss": 0.5014,
+      "step": 10677
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6921572770796952,
+      "learning_rate": 2.4504344935694996e-06,
+      "loss": 0.5261,
+      "step": 10678
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7199160789820862,
+      "learning_rate": 2.4495471506425183e-06,
+      "loss": 0.4785,
+      "step": 10679
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6071135882596962,
+      "learning_rate": 2.4486599162801674e-06,
+      "loss": 0.4319,
+      "step": 10680
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5683515544675386,
+      "learning_rate": 2.4477727905202085e-06,
+      "loss": 0.3997,
+      "step": 10681
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5807853982951335,
+      "learning_rate": 2.446885773400408e-06,
+      "loss": 0.4772,
+      "step": 10682
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6791670099245443,
+      "learning_rate": 2.4459988649585182e-06,
+      "loss": 0.4622,
+      "step": 10683
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6014193527221852,
+      "learning_rate": 2.445112065232299e-06,
+      "loss": 0.4926,
+      "step": 10684
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6813237596867474,
+      "learning_rate": 2.444225374259493e-06,
+      "loss": 0.4753,
+      "step": 10685
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8235857899859498,
+      "learning_rate": 2.4433387920778475e-06,
+      "loss": 0.5193,
+      "step": 10686
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7172004316474528,
+      "learning_rate": 2.4424523187250974e-06,
+      "loss": 0.4809,
+      "step": 10687
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.4542885268455246,
+      "learning_rate": 2.441565954238982e-06,
+      "loss": 0.4996,
+      "step": 10688
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7140273438417358,
+      "learning_rate": 2.4406796986572256e-06,
+      "loss": 0.5256,
+      "step": 10689
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5816882506093489,
+      "learning_rate": 2.439793552017559e-06,
+      "loss": 0.4375,
+      "step": 10690
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6688746931103752,
+      "learning_rate": 2.438907514357697e-06,
+      "loss": 0.4691,
+      "step": 10691
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7911759252836437,
+      "learning_rate": 2.43802158571536e-06,
+      "loss": 0.4808,
+      "step": 10692
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6997345922810209,
+      "learning_rate": 2.437135766128255e-06,
+      "loss": 0.441,
+      "step": 10693
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7697127989619299,
+      "learning_rate": 2.4362500556340936e-06,
+      "loss": 0.5065,
+      "step": 10694
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.76261356575478,
+      "learning_rate": 2.4353644542705703e-06,
+      "loss": 0.4966,
+      "step": 10695
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5992984772511316,
+      "learning_rate": 2.43447896207539e-06,
+      "loss": 0.5088,
+      "step": 10696
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7693345937281759,
+      "learning_rate": 2.4335935790862406e-06,
+      "loss": 0.5733,
+      "step": 10697
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7374141415594203,
+      "learning_rate": 2.432708305340814e-06,
+      "loss": 0.5348,
+      "step": 10698
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6752949821992592,
+      "learning_rate": 2.431823140876788e-06,
+      "loss": 0.4986,
+      "step": 10699
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5494342031098339,
+      "learning_rate": 2.430938085731847e-06,
+      "loss": 0.4069,
+      "step": 10700
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6345041032562396,
+      "learning_rate": 2.43005313994366e-06,
+      "loss": 0.4874,
+      "step": 10701
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7599971000903185,
+      "learning_rate": 2.4291683035499004e-06,
+      "loss": 0.5238,
+      "step": 10702
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7380009544892384,
+      "learning_rate": 2.4282835765882286e-06,
+      "loss": 0.5222,
+      "step": 10703
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7670832887018235,
+      "learning_rate": 2.4273989590963093e-06,
+      "loss": 0.5281,
+      "step": 10704
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5904091697497313,
+      "learning_rate": 2.426514451111793e-06,
+      "loss": 0.473,
+      "step": 10705
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6331370955412378,
+      "learning_rate": 2.4256300526723352e-06,
+      "loss": 0.4478,
+      "step": 10706
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6468386643533353,
+      "learning_rate": 2.4247457638155775e-06,
+      "loss": 0.4845,
+      "step": 10707
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6339230424769525,
+      "learning_rate": 2.4238615845791636e-06,
+      "loss": 0.5139,
+      "step": 10708
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8816016484557027,
+      "learning_rate": 2.4229775150007295e-06,
+      "loss": 0.5033,
+      "step": 10709
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6510093955007307,
+      "learning_rate": 2.422093555117909e-06,
+      "loss": 0.4568,
+      "step": 10710
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5889514282845364,
+      "learning_rate": 2.421209704968327e-06,
+      "loss": 0.4352,
+      "step": 10711
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6342082783199369,
+      "learning_rate": 2.420325964589609e-06,
+      "loss": 0.4202,
+      "step": 10712
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6272517155245021,
+      "learning_rate": 2.4194423340193696e-06,
+      "loss": 0.5029,
+      "step": 10713
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7838106229972467,
+      "learning_rate": 2.418558813295225e-06,
+      "loss": 0.484,
+      "step": 10714
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.939730119450188,
+      "learning_rate": 2.417675402454781e-06,
+      "loss": 0.5929,
+      "step": 10715
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.730871296229932,
+      "learning_rate": 2.416792101535645e-06,
+      "loss": 0.5235,
+      "step": 10716
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5420394027454228,
+      "learning_rate": 2.4159089105754123e-06,
+      "loss": 0.4323,
+      "step": 10717
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5699369533277572,
+      "learning_rate": 2.4150258296116825e-06,
+      "loss": 0.4385,
+      "step": 10718
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6008010964580018,
+      "learning_rate": 2.41414285868204e-06,
+      "loss": 0.456,
+      "step": 10719
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6243355450781678,
+      "learning_rate": 2.4132599978240727e-06,
+      "loss": 0.4591,
+      "step": 10720
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.0238150269553674,
+      "learning_rate": 2.4123772470753636e-06,
+      "loss": 0.5906,
+      "step": 10721
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7473902968138395,
+      "learning_rate": 2.4114946064734845e-06,
+      "loss": 0.4959,
+      "step": 10722
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6653613205142198,
+      "learning_rate": 2.4106120760560105e-06,
+      "loss": 0.49,
+      "step": 10723
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7572531889262955,
+      "learning_rate": 2.4097296558605034e-06,
+      "loss": 0.4649,
+      "step": 10724
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5732860034865839,
+      "learning_rate": 2.408847345924529e-06,
+      "loss": 0.4335,
+      "step": 10725
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5664822620728239,
+      "learning_rate": 2.407965146285642e-06,
+      "loss": 0.4193,
+      "step": 10726
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6099452083868128,
+      "learning_rate": 2.4070830569813954e-06,
+      "loss": 0.4613,
+      "step": 10727
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6531283756369385,
+      "learning_rate": 2.406201078049339e-06,
+      "loss": 0.4591,
+      "step": 10728
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6071837708400514,
+      "learning_rate": 2.4053192095270122e-06,
+      "loss": 0.4863,
+      "step": 10729
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.726671450738999,
+      "learning_rate": 2.4044374514519575e-06,
+      "loss": 0.5041,
+      "step": 10730
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6753116282711898,
+      "learning_rate": 2.403555803861704e-06,
+      "loss": 0.5246,
+      "step": 10731
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7845088803810617,
+      "learning_rate": 2.402674266793783e-06,
+      "loss": 0.5073,
+      "step": 10732
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5858019626031771,
+      "learning_rate": 2.4017928402857206e-06,
+      "loss": 0.4947,
+      "step": 10733
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7594251192336049,
+      "learning_rate": 2.400911524375032e-06,
+      "loss": 0.5059,
+      "step": 10734
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6436294907042814,
+      "learning_rate": 2.400030319099237e-06,
+      "loss": 0.4427,
+      "step": 10735
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.665587087036889,
+      "learning_rate": 2.3991492244958403e-06,
+      "loss": 0.4547,
+      "step": 10736
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7935765655517953,
+      "learning_rate": 2.398268240602352e-06,
+      "loss": 0.4977,
+      "step": 10737
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5973192112742913,
+      "learning_rate": 2.3973873674562693e-06,
+      "loss": 0.4672,
+      "step": 10738
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6393252440505999,
+      "learning_rate": 2.3965066050950913e-06,
+      "loss": 0.4601,
+      "step": 10739
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6370672488453737,
+      "learning_rate": 2.3956259535563054e-06,
+      "loss": 0.4552,
+      "step": 10740
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6482178678751401,
+      "learning_rate": 2.394745412877403e-06,
+      "loss": 0.4786,
+      "step": 10741
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6032793464429704,
+      "learning_rate": 2.3938649830958606e-06,
+      "loss": 0.5129,
+      "step": 10742
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8323148011681242,
+      "learning_rate": 2.3929846642491595e-06,
+      "loss": 0.4937,
+      "step": 10743
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5627431230471072,
+      "learning_rate": 2.3921044563747685e-06,
+      "loss": 0.428,
+      "step": 10744
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6393055905578594,
+      "learning_rate": 2.3912243595101574e-06,
+      "loss": 0.4671,
+      "step": 10745
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5901338716140343,
+      "learning_rate": 2.3903443736927897e-06,
+      "loss": 0.4859,
+      "step": 10746
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7812220224675036,
+      "learning_rate": 2.3894644989601236e-06,
+      "loss": 0.526,
+      "step": 10747
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5606310442516241,
+      "learning_rate": 2.38858473534961e-06,
+      "loss": 0.4667,
+      "step": 10748
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.693223459316275,
+      "learning_rate": 2.3877050828987016e-06,
+      "loss": 0.4727,
+      "step": 10749
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6675015607182109,
+      "learning_rate": 2.3868255416448377e-06,
+      "loss": 0.5087,
+      "step": 10750
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6886864857442677,
+      "learning_rate": 2.385946111625462e-06,
+      "loss": 0.46,
+      "step": 10751
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6102865259761342,
+      "learning_rate": 2.385066792878005e-06,
+      "loss": 0.4672,
+      "step": 10752
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7018802541894525,
+      "learning_rate": 2.384187585439901e-06,
+      "loss": 0.4813,
+      "step": 10753
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6890841022637195,
+      "learning_rate": 2.3833084893485705e-06,
+      "loss": 0.4756,
+      "step": 10754
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7642755706983619,
+      "learning_rate": 2.382429504641437e-06,
+      "loss": 0.4847,
+      "step": 10755
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8206867294764637,
+      "learning_rate": 2.381550631355914e-06,
+      "loss": 0.469,
+      "step": 10756
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.058137301693994,
+      "learning_rate": 2.3806718695294124e-06,
+      "loss": 0.4359,
+      "step": 10757
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7446157661237356,
+      "learning_rate": 2.3797932191993396e-06,
+      "loss": 0.4761,
+      "step": 10758
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7050987456196054,
+      "learning_rate": 2.3789146804030984e-06,
+      "loss": 0.4839,
+      "step": 10759
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7544500047765144,
+      "learning_rate": 2.3780362531780815e-06,
+      "loss": 0.509,
+      "step": 10760
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8407991862856738,
+      "learning_rate": 2.377157937561684e-06,
+      "loss": 0.5663,
+      "step": 10761
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7558720128445902,
+      "learning_rate": 2.376279733591291e-06,
+      "loss": 0.4997,
+      "step": 10762
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6258939976241674,
+      "learning_rate": 2.375401641304287e-06,
+      "loss": 0.4836,
+      "step": 10763
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7867419421900359,
+      "learning_rate": 2.374523660738046e-06,
+      "loss": 0.5363,
+      "step": 10764
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7395039651205694,
+      "learning_rate": 2.3736457919299446e-06,
+      "loss": 0.4981,
+      "step": 10765
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7405534995697715,
+      "learning_rate": 2.3727680349173475e-06,
+      "loss": 0.5776,
+      "step": 10766
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7104559397257032,
+      "learning_rate": 2.3718903897376216e-06,
+      "loss": 0.4762,
+      "step": 10767
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6542246828416545,
+      "learning_rate": 2.371012856428122e-06,
+      "loss": 0.4525,
+      "step": 10768
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5609690670930079,
+      "learning_rate": 2.370135435026206e-06,
+      "loss": 0.4263,
+      "step": 10769
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8051835397134139,
+      "learning_rate": 2.3692581255692167e-06,
+      "loss": 0.5298,
+      "step": 10770
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7743788436099871,
+      "learning_rate": 2.368380928094506e-06,
+      "loss": 0.5792,
+      "step": 10771
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6776847729356127,
+      "learning_rate": 2.3675038426394078e-06,
+      "loss": 0.5239,
+      "step": 10772
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6080921980201661,
+      "learning_rate": 2.3666268692412603e-06,
+      "loss": 0.4805,
+      "step": 10773
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6871087491999675,
+      "learning_rate": 2.365750007937391e-06,
+      "loss": 0.4787,
+      "step": 10774
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7243648822500646,
+      "learning_rate": 2.3648732587651268e-06,
+      "loss": 0.5107,
+      "step": 10775
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6034248568715802,
+      "learning_rate": 2.3639966217617868e-06,
+      "loss": 0.5041,
+      "step": 10776
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.0120335372650437,
+      "learning_rate": 2.363120096964688e-06,
+      "loss": 0.5738,
+      "step": 10777
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7415451815977266,
+      "learning_rate": 2.3622436844111384e-06,
+      "loss": 0.556,
+      "step": 10778
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8256111580997718,
+      "learning_rate": 2.361367384138448e-06,
+      "loss": 0.5658,
+      "step": 10779
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.719510323500556,
+      "learning_rate": 2.360491196183915e-06,
+      "loss": 0.5292,
+      "step": 10780
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.004692350210955,
+      "learning_rate": 2.359615120584839e-06,
+      "loss": 0.559,
+      "step": 10781
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7024380001621336,
+      "learning_rate": 2.3587391573785073e-06,
+      "loss": 0.533,
+      "step": 10782
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5796511462605038,
+      "learning_rate": 2.35786330660221e-06,
+      "loss": 0.4513,
+      "step": 10783
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8258690913604891,
+      "learning_rate": 2.3569875682932304e-06,
+      "loss": 0.5415,
+      "step": 10784
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6129396432400224,
+      "learning_rate": 2.356111942488842e-06,
+      "loss": 0.5035,
+      "step": 10785
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7709876421865782,
+      "learning_rate": 2.3552364292263215e-06,
+      "loss": 0.6035,
+      "step": 10786
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7091236168630716,
+      "learning_rate": 2.3543610285429332e-06,
+      "loss": 0.4916,
+      "step": 10787
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5715637774864741,
+      "learning_rate": 2.3534857404759414e-06,
+      "loss": 0.478,
+      "step": 10788
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6525523934717318,
+      "learning_rate": 2.352610565062606e-06,
+      "loss": 0.4162,
+      "step": 10789
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6228413227067008,
+      "learning_rate": 2.351735502340178e-06,
+      "loss": 0.5088,
+      "step": 10790
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.729817446892084,
+      "learning_rate": 2.350860552345908e-06,
+      "loss": 0.5746,
+      "step": 10791
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6946337814457026,
+      "learning_rate": 2.349985715117038e-06,
+      "loss": 0.5064,
+      "step": 10792
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5995985235027571,
+      "learning_rate": 2.3491109906908095e-06,
+      "loss": 0.4355,
+      "step": 10793
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6098114528755813,
+      "learning_rate": 2.348236379104453e-06,
+      "loss": 0.453,
+      "step": 10794
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5293180168757409,
+      "learning_rate": 2.3473618803951997e-06,
+      "loss": 0.4282,
+      "step": 10795
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6461010280129098,
+      "learning_rate": 2.3464874946002762e-06,
+      "loss": 0.4835,
+      "step": 10796
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.604742512563059,
+      "learning_rate": 2.3456132217568994e-06,
+      "loss": 0.4334,
+      "step": 10797
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6728208997366567,
+      "learning_rate": 2.3447390619022873e-06,
+      "loss": 0.4376,
+      "step": 10798
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6260716571252182,
+      "learning_rate": 2.3438650150736465e-06,
+      "loss": 0.438,
+      "step": 10799
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6411162499271806,
+      "learning_rate": 2.342991081308186e-06,
+      "loss": 0.5091,
+      "step": 10800
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7894618104541057,
+      "learning_rate": 2.342117260643103e-06,
+      "loss": 0.4988,
+      "step": 10801
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7268751297462787,
+      "learning_rate": 2.3412435531155963e-06,
+      "loss": 0.4491,
+      "step": 10802
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8370397054839325,
+      "learning_rate": 2.3403699587628537e-06,
+      "loss": 0.507,
+      "step": 10803
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6221575888969254,
+      "learning_rate": 2.339496477622065e-06,
+      "loss": 0.4539,
+      "step": 10804
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7669903378821644,
+      "learning_rate": 2.3386231097304073e-06,
+      "loss": 0.5337,
+      "step": 10805
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7556400172219717,
+      "learning_rate": 2.3377498551250612e-06,
+      "loss": 0.5698,
+      "step": 10806
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5907499487355397,
+      "learning_rate": 2.3368767138431926e-06,
+      "loss": 0.4793,
+      "step": 10807
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.9718417545714607,
+      "learning_rate": 2.3360036859219758e-06,
+      "loss": 0.5496,
+      "step": 10808
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.80538505610934,
+      "learning_rate": 2.3351307713985678e-06,
+      "loss": 0.4847,
+      "step": 10809
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6634063433384094,
+      "learning_rate": 2.334257970310129e-06,
+      "loss": 0.4551,
+      "step": 10810
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7258577887381235,
+      "learning_rate": 2.3333852826938082e-06,
+      "loss": 0.5441,
+      "step": 10811
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6722498850302644,
+      "learning_rate": 2.3325127085867554e-06,
+      "loss": 0.4432,
+      "step": 10812
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7829308135547088,
+      "learning_rate": 2.331640248026112e-06,
+      "loss": 0.4876,
+      "step": 10813
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7334023455009091,
+      "learning_rate": 2.3307679010490175e-06,
+      "loss": 0.5017,
+      "step": 10814
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6371490893324795,
+      "learning_rate": 2.3298956676926023e-06,
+      "loss": 0.4521,
+      "step": 10815
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6348561939785046,
+      "learning_rate": 2.329023547993998e-06,
+      "loss": 0.453,
+      "step": 10816
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7845462444401949,
+      "learning_rate": 2.3281515419903233e-06,
+      "loss": 0.4941,
+      "step": 10817
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7707273584196197,
+      "learning_rate": 2.3272796497187023e-06,
+      "loss": 0.5087,
+      "step": 10818
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.691895113348418,
+      "learning_rate": 2.3264078712162413e-06,
+      "loss": 0.4724,
+      "step": 10819
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6067262766921584,
+      "learning_rate": 2.325536206520058e-06,
+      "loss": 0.442,
+      "step": 10820
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6007288267040728,
+      "learning_rate": 2.32466465566725e-06,
+      "loss": 0.4777,
+      "step": 10821
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8321389327862625,
+      "learning_rate": 2.32379321869492e-06,
+      "loss": 0.5049,
+      "step": 10822
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7320317635548815,
+      "learning_rate": 2.3229218956401593e-06,
+      "loss": 0.5155,
+      "step": 10823
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6482921480648263,
+      "learning_rate": 2.3220506865400604e-06,
+      "loss": 0.4684,
+      "step": 10824
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8111920894798678,
+      "learning_rate": 2.321179591431705e-06,
+      "loss": 0.4723,
+      "step": 10825
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6701617418798336,
+      "learning_rate": 2.3203086103521753e-06,
+      "loss": 0.4953,
+      "step": 10826
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8428668487849036,
+      "learning_rate": 2.3194377433385434e-06,
+      "loss": 0.5295,
+      "step": 10827
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.727807295443101,
+      "learning_rate": 2.318566990427883e-06,
+      "loss": 0.5495,
+      "step": 10828
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6395711581650799,
+      "learning_rate": 2.3176963516572546e-06,
+      "loss": 0.4289,
+      "step": 10829
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6438211565663589,
+      "learning_rate": 2.3168258270637233e-06,
+      "loss": 0.4228,
+      "step": 10830
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.618310023751601,
+      "learning_rate": 2.3159554166843408e-06,
+      "loss": 0.4546,
+      "step": 10831
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7663220610981202,
+      "learning_rate": 2.315085120556159e-06,
+      "loss": 0.5122,
+      "step": 10832
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5277908735189812,
+      "learning_rate": 2.314214938716224e-06,
+      "loss": 0.4127,
+      "step": 10833
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6815308361486874,
+      "learning_rate": 2.3133448712015784e-06,
+      "loss": 0.4891,
+      "step": 10834
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6643055000468925,
+      "learning_rate": 2.312474918049254e-06,
+      "loss": 0.4648,
+      "step": 10835
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7628619037188938,
+      "learning_rate": 2.311605079296286e-06,
+      "loss": 0.5443,
+      "step": 10836
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.697087369735636,
+      "learning_rate": 2.310735354979698e-06,
+      "loss": 0.4966,
+      "step": 10837
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.63360614192725,
+      "learning_rate": 2.309865745136513e-06,
+      "loss": 0.4964,
+      "step": 10838
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5240209848865982,
+      "learning_rate": 2.3089962498037454e-06,
+      "loss": 0.386,
+      "step": 10839
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.784939710411291,
+      "learning_rate": 2.3081268690184106e-06,
+      "loss": 0.4941,
+      "step": 10840
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8528359232804006,
+      "learning_rate": 2.307257602817511e-06,
+      "loss": 0.5312,
+      "step": 10841
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8980076187595706,
+      "learning_rate": 2.306388451238052e-06,
+      "loss": 0.5149,
+      "step": 10842
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6039802533407461,
+      "learning_rate": 2.3055194143170268e-06,
+      "loss": 0.4484,
+      "step": 10843
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.673218966898378,
+      "learning_rate": 2.30465049209143e-06,
+      "loss": 0.5021,
+      "step": 10844
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.71350439495478,
+      "learning_rate": 2.3037816845982512e-06,
+      "loss": 0.5002,
+      "step": 10845
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6059695953100277,
+      "learning_rate": 2.302912991874468e-06,
+      "loss": 0.4107,
+      "step": 10846
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.629674748262165,
+      "learning_rate": 2.302044413957062e-06,
+      "loss": 0.5115,
+      "step": 10847
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6957572241525941,
+      "learning_rate": 2.3011759508830017e-06,
+      "loss": 0.5358,
+      "step": 10848
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7079104543439373,
+      "learning_rate": 2.300307602689259e-06,
+      "loss": 0.5064,
+      "step": 10849
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6748084773974228,
+      "learning_rate": 2.2994393694127925e-06,
+      "loss": 0.4545,
+      "step": 10850
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8198493326666778,
+      "learning_rate": 2.298571251090563e-06,
+      "loss": 0.5341,
+      "step": 10851
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8354673963447915,
+      "learning_rate": 2.297703247759524e-06,
+      "loss": 0.485,
+      "step": 10852
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.609875454634327,
+      "learning_rate": 2.2968353594566204e-06,
+      "loss": 0.4674,
+      "step": 10853
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8023069330239143,
+      "learning_rate": 2.2959675862188e-06,
+      "loss": 0.5317,
+      "step": 10854
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.646843375110653,
+      "learning_rate": 2.2950999280829965e-06,
+      "loss": 0.4665,
+      "step": 10855
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8594025282298801,
+      "learning_rate": 2.294232385086146e-06,
+      "loss": 0.5593,
+      "step": 10856
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6386188645289982,
+      "learning_rate": 2.2933649572651774e-06,
+      "loss": 0.4471,
+      "step": 10857
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.595628600036395,
+      "learning_rate": 2.2924976446570117e-06,
+      "loss": 0.434,
+      "step": 10858
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6302137959335479,
+      "learning_rate": 2.2916304472985717e-06,
+      "loss": 0.4214,
+      "step": 10859
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7256589824548828,
+      "learning_rate": 2.2907633652267665e-06,
+      "loss": 0.4948,
+      "step": 10860
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5784285660717894,
+      "learning_rate": 2.2898963984785094e-06,
+      "loss": 0.4217,
+      "step": 10861
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6399792271673397,
+      "learning_rate": 2.2890295470907004e-06,
+      "loss": 0.5031,
+      "step": 10862
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8377256946518645,
+      "learning_rate": 2.288162811100243e-06,
+      "loss": 0.5681,
+      "step": 10863
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7997655840390412,
+      "learning_rate": 2.287296190544026e-06,
+      "loss": 0.5771,
+      "step": 10864
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6840156624585454,
+      "learning_rate": 2.2864296854589447e-06,
+      "loss": 0.5083,
+      "step": 10865
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7734226611635323,
+      "learning_rate": 2.2855632958818775e-06,
+      "loss": 0.5284,
+      "step": 10866
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7179434223053618,
+      "learning_rate": 2.2846970218497087e-06,
+      "loss": 0.5535,
+      "step": 10867
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.853643224698523,
+      "learning_rate": 2.2838308633993094e-06,
+      "loss": 0.562,
+      "step": 10868
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7325780615361955,
+      "learning_rate": 2.2829648205675505e-06,
+      "loss": 0.5552,
+      "step": 10869
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7512348455354795,
+      "learning_rate": 2.282098893391297e-06,
+      "loss": 0.5048,
+      "step": 10870
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6871867274601527,
+      "learning_rate": 2.2812330819074098e-06,
+      "loss": 0.4601,
+      "step": 10871
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6898907884307547,
+      "learning_rate": 2.2803673861527402e-06,
+      "loss": 0.4737,
+      "step": 10872
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7204682129670187,
+      "learning_rate": 2.2795018061641426e-06,
+      "loss": 0.4918,
+      "step": 10873
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7429270747384472,
+      "learning_rate": 2.278636341978458e-06,
+      "loss": 0.5492,
+      "step": 10874
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7478198372537292,
+      "learning_rate": 2.27777099363253e-06,
+      "loss": 0.4986,
+      "step": 10875
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6253559315027629,
+      "learning_rate": 2.2769057611631896e-06,
+      "loss": 0.4849,
+      "step": 10876
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7917345277229931,
+      "learning_rate": 2.276040644607271e-06,
+      "loss": 0.539,
+      "step": 10877
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.0129410880127359,
+      "learning_rate": 2.275175644001596e-06,
+      "loss": 0.529,
+      "step": 10878
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6416267711648092,
+      "learning_rate": 2.2743107593829882e-06,
+      "loss": 0.5089,
+      "step": 10879
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7531020512628711,
+      "learning_rate": 2.2734459907882593e-06,
+      "loss": 0.5236,
+      "step": 10880
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6191266285390059,
+      "learning_rate": 2.272581338254222e-06,
+      "loss": 0.4809,
+      "step": 10881
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7061625663095266,
+      "learning_rate": 2.2717168018176816e-06,
+      "loss": 0.4906,
+      "step": 10882
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6615318409787045,
+      "learning_rate": 2.270852381515441e-06,
+      "loss": 0.4588,
+      "step": 10883
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5958957323794096,
+      "learning_rate": 2.269988077384291e-06,
+      "loss": 0.4467,
+      "step": 10884
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7140301508599692,
+      "learning_rate": 2.2691238894610262e-06,
+      "loss": 0.5212,
+      "step": 10885
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.70847909896908,
+      "learning_rate": 2.2682598177824295e-06,
+      "loss": 0.4742,
+      "step": 10886
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7408743619337694,
+      "learning_rate": 2.2673958623852845e-06,
+      "loss": 0.5357,
+      "step": 10887
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.624318605854862,
+      "learning_rate": 2.266532023306364e-06,
+      "loss": 0.4861,
+      "step": 10888
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6330453646878533,
+      "learning_rate": 2.2656683005824414e-06,
+      "loss": 0.4498,
+      "step": 10889
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.830984947084523,
+      "learning_rate": 2.26480469425028e-06,
+      "loss": 0.5654,
+      "step": 10890
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.1469468777747078,
+      "learning_rate": 2.263941204346645e-06,
+      "loss": 0.5465,
+      "step": 10891
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7971891625506593,
+      "learning_rate": 2.263077830908287e-06,
+      "loss": 0.5525,
+      "step": 10892
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7687905450681011,
+      "learning_rate": 2.2622145739719618e-06,
+      "loss": 0.5422,
+      "step": 10893
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7859554968939656,
+      "learning_rate": 2.26135143357441e-06,
+      "loss": 0.5408,
+      "step": 10894
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7897704926868802,
+      "learning_rate": 2.26048840975238e-06,
+      "loss": 0.5684,
+      "step": 10895
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6984672281073324,
+      "learning_rate": 2.2596255025426016e-06,
+      "loss": 0.557,
+      "step": 10896
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7968758589073045,
+      "learning_rate": 2.2587627119818114e-06,
+      "loss": 0.5242,
+      "step": 10897
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7185412552560613,
+      "learning_rate": 2.25790003810673e-06,
+      "loss": 0.4765,
+      "step": 10898
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6079780037116617,
+      "learning_rate": 2.257037480954084e-06,
+      "loss": 0.4408,
+      "step": 10899
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6468975121598799,
+      "learning_rate": 2.2561750405605852e-06,
+      "loss": 0.4928,
+      "step": 10900
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7641646389324386,
+      "learning_rate": 2.2553127169629483e-06,
+      "loss": 0.4848,
+      "step": 10901
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6023575145887837,
+      "learning_rate": 2.2544505101978775e-06,
+      "loss": 0.4138,
+      "step": 10902
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6652981432363783,
+      "learning_rate": 2.253588420302076e-06,
+      "loss": 0.4721,
+      "step": 10903
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5903932658563311,
+      "learning_rate": 2.2527264473122375e-06,
+      "loss": 0.4271,
+      "step": 10904
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5738573114569012,
+      "learning_rate": 2.2518645912650565e-06,
+      "loss": 0.4052,
+      "step": 10905
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6396189298051217,
+      "learning_rate": 2.251002852197217e-06,
+      "loss": 0.4674,
+      "step": 10906
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8670799584371012,
+      "learning_rate": 2.2501412301454014e-06,
+      "loss": 0.5947,
+      "step": 10907
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8304689547212774,
+      "learning_rate": 2.249279725146288e-06,
+      "loss": 0.4901,
+      "step": 10908
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.776941287475717,
+      "learning_rate": 2.248418337236545e-06,
+      "loss": 0.4466,
+      "step": 10909
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6066677428957276,
+      "learning_rate": 2.247557066452842e-06,
+      "loss": 0.4477,
+      "step": 10910
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5851247005247016,
+      "learning_rate": 2.246695912831837e-06,
+      "loss": 0.4324,
+      "step": 10911
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6741998445693816,
+      "learning_rate": 2.2458348764101893e-06,
+      "loss": 0.5196,
+      "step": 10912
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6836705911603268,
+      "learning_rate": 2.2449739572245522e-06,
+      "loss": 0.4922,
+      "step": 10913
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5734677035468869,
+      "learning_rate": 2.244113155311567e-06,
+      "loss": 0.4967,
+      "step": 10914
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6037662181064927,
+      "learning_rate": 2.2432524707078806e-06,
+      "loss": 0.4282,
+      "step": 10915
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6634103154685594,
+      "learning_rate": 2.242391903450126e-06,
+      "loss": 0.511,
+      "step": 10916
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6992329402401665,
+      "learning_rate": 2.2415314535749364e-06,
+      "loss": 0.5075,
+      "step": 10917
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5164375626469455,
+      "learning_rate": 2.2406711211189376e-06,
+      "loss": 0.4375,
+      "step": 10918
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7710943939776963,
+      "learning_rate": 2.2398109061187507e-06,
+      "loss": 0.5269,
+      "step": 10919
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7076536057752935,
+      "learning_rate": 2.2389508086109953e-06,
+      "loss": 0.5515,
+      "step": 10920
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5653953532490982,
+      "learning_rate": 2.2380908286322796e-06,
+      "loss": 0.498,
+      "step": 10921
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7356905916586278,
+      "learning_rate": 2.237230966219213e-06,
+      "loss": 0.4974,
+      "step": 10922
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6122952244606756,
+      "learning_rate": 2.236371221408393e-06,
+      "loss": 0.4574,
+      "step": 10923
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.6170098335556904,
+      "learning_rate": 2.2355115942364212e-06,
+      "loss": 0.5068,
+      "step": 10924
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.716916126601903,
+      "learning_rate": 2.2346520847398856e-06,
+      "loss": 0.5341,
+      "step": 10925
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7358376455793131,
+      "learning_rate": 2.233792692955375e-06,
+      "loss": 0.4922,
+      "step": 10926
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7445956345088092,
+      "learning_rate": 2.2329334189194684e-06,
+      "loss": 0.484,
+      "step": 10927
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7791986312983723,
+      "learning_rate": 2.232074262668746e-06,
+      "loss": 0.5081,
+      "step": 10928
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6554900309954337,
+      "learning_rate": 2.2312152242397748e-06,
+      "loss": 0.5061,
+      "step": 10929
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.9264581091645545,
+      "learning_rate": 2.2303563036691262e-06,
+      "loss": 0.5572,
+      "step": 10930
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7857140826550714,
+      "learning_rate": 2.229497500993355e-06,
+      "loss": 0.4902,
+      "step": 10931
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6708662137603609,
+      "learning_rate": 2.228638816249026e-06,
+      "loss": 0.4536,
+      "step": 10932
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7342422332790777,
+      "learning_rate": 2.227780249472684e-06,
+      "loss": 0.4723,
+      "step": 10933
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5941908653503654,
+      "learning_rate": 2.226921800700881e-06,
+      "loss": 0.4013,
+      "step": 10934
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5758168265318097,
+      "learning_rate": 2.2260634699701527e-06,
+      "loss": 0.4454,
+      "step": 10935
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6809199870318094,
+      "learning_rate": 2.225205257317041e-06,
+      "loss": 0.4624,
+      "step": 10936
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7105686784561956,
+      "learning_rate": 2.2243471627780716e-06,
+      "loss": 0.491,
+      "step": 10937
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6869496746710227,
+      "learning_rate": 2.223489186389776e-06,
+      "loss": 0.4574,
+      "step": 10938
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6150084343203703,
+      "learning_rate": 2.2226313281886718e-06,
+      "loss": 0.4991,
+      "step": 10939
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6877291962032914,
+      "learning_rate": 2.2217735882112784e-06,
+      "loss": 0.4972,
+      "step": 10940
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8457120854417003,
+      "learning_rate": 2.2209159664941034e-06,
+      "loss": 0.45,
+      "step": 10941
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6081577001832593,
+      "learning_rate": 2.220058463073657e-06,
+      "loss": 0.4321,
+      "step": 10942
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.624573766215268,
+      "learning_rate": 2.219201077986435e-06,
+      "loss": 0.4753,
+      "step": 10943
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7625434569000092,
+      "learning_rate": 2.21834381126894e-06,
+      "loss": 0.5031,
+      "step": 10944
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7393329927868117,
+      "learning_rate": 2.2174866629576585e-06,
+      "loss": 0.5064,
+      "step": 10945
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6317837046751107,
+      "learning_rate": 2.2166296330890802e-06,
+      "loss": 0.4425,
+      "step": 10946
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6471990765402897,
+      "learning_rate": 2.2157727216996823e-06,
+      "loss": 0.4404,
+      "step": 10947
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6592984573783606,
+      "learning_rate": 2.2149159288259435e-06,
+      "loss": 0.4622,
+      "step": 10948
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.541674743752519,
+      "learning_rate": 2.2140592545043326e-06,
+      "loss": 0.4044,
+      "step": 10949
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6076175752250428,
+      "learning_rate": 2.2132026987713185e-06,
+      "loss": 0.457,
+      "step": 10950
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5562338123275646,
+      "learning_rate": 2.2123462616633585e-06,
+      "loss": 0.451,
+      "step": 10951
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6705609066312985,
+      "learning_rate": 2.211489943216911e-06,
+      "loss": 0.5134,
+      "step": 10952
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7642569088347329,
+      "learning_rate": 2.210633743468425e-06,
+      "loss": 0.5315,
+      "step": 10953
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.841355749038376,
+      "learning_rate": 2.209777662454348e-06,
+      "loss": 0.5276,
+      "step": 10954
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0077373634766211,
+      "learning_rate": 2.2089217002111184e-06,
+      "loss": 0.5449,
+      "step": 10955
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.584543446062315,
+      "learning_rate": 2.2080658567751727e-06,
+      "loss": 0.5077,
+      "step": 10956
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7767332844256136,
+      "learning_rate": 2.207210132182942e-06,
+      "loss": 0.5309,
+      "step": 10957
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6926558720085354,
+      "learning_rate": 2.2063545264708535e-06,
+      "loss": 0.479,
+      "step": 10958
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6884824516269165,
+      "learning_rate": 2.2054990396753236e-06,
+      "loss": 0.4546,
+      "step": 10959
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6168371490437435,
+      "learning_rate": 2.204643671832771e-06,
+      "loss": 0.4368,
+      "step": 10960
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7702975252598642,
+      "learning_rate": 2.203788422979603e-06,
+      "loss": 0.5454,
+      "step": 10961
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8119574575472767,
+      "learning_rate": 2.2029332931522284e-06,
+      "loss": 0.5198,
+      "step": 10962
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.9934127618301879,
+      "learning_rate": 2.202078282387044e-06,
+      "loss": 0.4838,
+      "step": 10963
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6712955357044467,
+      "learning_rate": 2.2012233907204474e-06,
+      "loss": 0.4438,
+      "step": 10964
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5654238662769213,
+      "learning_rate": 2.2003686181888257e-06,
+      "loss": 0.4054,
+      "step": 10965
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.60106904253335,
+      "learning_rate": 2.1995139648285673e-06,
+      "loss": 0.4268,
+      "step": 10966
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6669297735871007,
+      "learning_rate": 2.198659430676049e-06,
+      "loss": 0.4651,
+      "step": 10967
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7193298588316885,
+      "learning_rate": 2.1978050157676465e-06,
+      "loss": 0.4746,
+      "step": 10968
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6321282919246608,
+      "learning_rate": 2.1969507201397323e-06,
+      "loss": 0.4319,
+      "step": 10969
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8279561336571408,
+      "learning_rate": 2.196096543828666e-06,
+      "loss": 0.5255,
+      "step": 10970
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8142394644349056,
+      "learning_rate": 2.195242486870812e-06,
+      "loss": 0.4659,
+      "step": 10971
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7207852920771451,
+      "learning_rate": 2.194388549302521e-06,
+      "loss": 0.4924,
+      "step": 10972
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6702533859239594,
+      "learning_rate": 2.193534731160144e-06,
+      "loss": 0.5134,
+      "step": 10973
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7363740120200745,
+      "learning_rate": 2.1926810324800275e-06,
+      "loss": 0.4671,
+      "step": 10974
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6283988046886836,
+      "learning_rate": 2.191827453298506e-06,
+      "loss": 0.4751,
+      "step": 10975
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6227413034732902,
+      "learning_rate": 2.190973993651918e-06,
+      "loss": 0.4862,
+      "step": 10976
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5495609392943195,
+      "learning_rate": 2.1901206535765896e-06,
+      "loss": 0.4458,
+      "step": 10977
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6314091810114084,
+      "learning_rate": 2.189267433108847e-06,
+      "loss": 0.4807,
+      "step": 10978
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7020886683319169,
+      "learning_rate": 2.1884143322850067e-06,
+      "loss": 0.4894,
+      "step": 10979
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6301172822728136,
+      "learning_rate": 2.1875613511413835e-06,
+      "loss": 0.4169,
+      "step": 10980
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6803946015984277,
+      "learning_rate": 2.1867084897142876e-06,
+      "loss": 0.5174,
+      "step": 10981
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8329509918098765,
+      "learning_rate": 2.18585574804002e-06,
+      "loss": 0.5136,
+      "step": 10982
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6443847846269465,
+      "learning_rate": 2.1850031261548825e-06,
+      "loss": 0.4595,
+      "step": 10983
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5957642063948049,
+      "learning_rate": 2.1841506240951644e-06,
+      "loss": 0.4559,
+      "step": 10984
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7878583552992771,
+      "learning_rate": 2.183298241897158e-06,
+      "loss": 0.4749,
+      "step": 10985
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6979600090124072,
+      "learning_rate": 2.1824459795971426e-06,
+      "loss": 0.4922,
+      "step": 10986
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8209454940473095,
+      "learning_rate": 2.1815938372314e-06,
+      "loss": 0.5243,
+      "step": 10987
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6023625779386415,
+      "learning_rate": 2.1807418148362e-06,
+      "loss": 0.4576,
+      "step": 10988
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.615940588367025,
+      "learning_rate": 2.1798899124478135e-06,
+      "loss": 0.4706,
+      "step": 10989
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.61390045257212,
+      "learning_rate": 2.1790381301024993e-06,
+      "loss": 0.4279,
+      "step": 10990
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.555138987211361,
+      "learning_rate": 2.17818646783652e-06,
+      "loss": 0.4089,
+      "step": 10991
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5683507604892607,
+      "learning_rate": 2.177334925686123e-06,
+      "loss": 0.499,
+      "step": 10992
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7343151709091383,
+      "learning_rate": 2.1764835036875585e-06,
+      "loss": 0.4961,
+      "step": 10993
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7442754463906057,
+      "learning_rate": 2.1756322018770685e-06,
+      "loss": 0.481,
+      "step": 10994
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5611698795914328,
+      "learning_rate": 2.1747810202908924e-06,
+      "loss": 0.4069,
+      "step": 10995
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6391799824310542,
+      "learning_rate": 2.1739299589652585e-06,
+      "loss": 0.4793,
+      "step": 10996
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0105235284549798,
+      "learning_rate": 2.173079017936397e-06,
+      "loss": 0.5128,
+      "step": 10997
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6317468501220947,
+      "learning_rate": 2.1722281972405263e-06,
+      "loss": 0.4406,
+      "step": 10998
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5781795300100511,
+      "learning_rate": 2.171377496913867e-06,
+      "loss": 0.4915,
+      "step": 10999
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8039107377411533,
+      "learning_rate": 2.1705269169926265e-06,
+      "loss": 0.5057,
+      "step": 11000
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.647062086882824,
+      "learning_rate": 2.1696764575130155e-06,
+      "loss": 0.4705,
+      "step": 11001
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6032750648057195,
+      "learning_rate": 2.1688261185112314e-06,
+      "loss": 0.5155,
+      "step": 11002
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8725981161136266,
+      "learning_rate": 2.167975900023474e-06,
+      "loss": 0.5371,
+      "step": 11003
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7612226804291636,
+      "learning_rate": 2.167125802085931e-06,
+      "loss": 0.5239,
+      "step": 11004
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7044912290668476,
+      "learning_rate": 2.16627582473479e-06,
+      "loss": 0.5128,
+      "step": 11005
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6814329210257032,
+      "learning_rate": 2.165425968006232e-06,
+      "loss": 0.4399,
+      "step": 11006
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6221781101872258,
+      "learning_rate": 2.1645762319364346e-06,
+      "loss": 0.4438,
+      "step": 11007
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7546992385737619,
+      "learning_rate": 2.163726616561564e-06,
+      "loss": 0.4564,
+      "step": 11008
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6794415969231482,
+      "learning_rate": 2.16287712191779e-06,
+      "loss": 0.4865,
+      "step": 11009
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6438880633732903,
+      "learning_rate": 2.1620277480412684e-06,
+      "loss": 0.473,
+      "step": 11010
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7669935149515065,
+      "learning_rate": 2.1611784949681585e-06,
+      "loss": 0.5343,
+      "step": 11011
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8208067960589795,
+      "learning_rate": 2.1603293627346074e-06,
+      "loss": 0.5343,
+      "step": 11012
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5818989720137931,
+      "learning_rate": 2.1594803513767626e-06,
+      "loss": 0.4405,
+      "step": 11013
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7124391133639624,
+      "learning_rate": 2.15863146093076e-06,
+      "loss": 0.4811,
+      "step": 11014
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6852734525096708,
+      "learning_rate": 2.157782691432739e-06,
+      "loss": 0.5071,
+      "step": 11015
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6114566431044908,
+      "learning_rate": 2.156934042918824e-06,
+      "loss": 0.4313,
+      "step": 11016
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6343984270692141,
+      "learning_rate": 2.1560855154251433e-06,
+      "loss": 0.4384,
+      "step": 11017
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8647573277877423,
+      "learning_rate": 2.155237108987811e-06,
+      "loss": 0.4109,
+      "step": 11018
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6361098323114568,
+      "learning_rate": 2.1543888236429483e-06,
+      "loss": 0.4358,
+      "step": 11019
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7087256061379335,
+      "learning_rate": 2.1535406594266577e-06,
+      "loss": 0.4257,
+      "step": 11020
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8394145324181174,
+      "learning_rate": 2.1526926163750476e-06,
+      "loss": 0.4702,
+      "step": 11021
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6080438234027822,
+      "learning_rate": 2.151844694524212e-06,
+      "loss": 0.4134,
+      "step": 11022
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6440188023966902,
+      "learning_rate": 2.150996893910248e-06,
+      "loss": 0.4978,
+      "step": 11023
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8782516481647507,
+      "learning_rate": 2.150149214569241e-06,
+      "loss": 0.5174,
+      "step": 11024
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5621256069500332,
+      "learning_rate": 2.1493016565372764e-06,
+      "loss": 0.4472,
+      "step": 11025
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5990425563435783,
+      "learning_rate": 2.1484542198504294e-06,
+      "loss": 0.4666,
+      "step": 11026
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7703837477702888,
+      "learning_rate": 2.1476069045447756e-06,
+      "loss": 0.5427,
+      "step": 11027
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6400067325818959,
+      "learning_rate": 2.1467597106563786e-06,
+      "loss": 0.4565,
+      "step": 11028
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6537877982122922,
+      "learning_rate": 2.145912638221305e-06,
+      "loss": 0.5026,
+      "step": 11029
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7826774851046876,
+      "learning_rate": 2.1450656872756086e-06,
+      "loss": 0.5036,
+      "step": 11030
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6172469896270728,
+      "learning_rate": 2.144218857855343e-06,
+      "loss": 0.4702,
+      "step": 11031
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6895217052906698,
+      "learning_rate": 2.1433721499965565e-06,
+      "loss": 0.4968,
+      "step": 11032
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6139194578998511,
+      "learning_rate": 2.142525563735287e-06,
+      "loss": 0.4577,
+      "step": 11033
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7523131711554061,
+      "learning_rate": 2.141679099107575e-06,
+      "loss": 0.5359,
+      "step": 11034
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8977456841423956,
+      "learning_rate": 2.1408327561494485e-06,
+      "loss": 0.5379,
+      "step": 11035
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5891850449584672,
+      "learning_rate": 2.1399865348969345e-06,
+      "loss": 0.4571,
+      "step": 11036
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5442590572440746,
+      "learning_rate": 2.1391404353860562e-06,
+      "loss": 0.4469,
+      "step": 11037
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7141649666533418,
+      "learning_rate": 2.1382944576528257e-06,
+      "loss": 0.5459,
+      "step": 11038
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7935313665736218,
+      "learning_rate": 2.137448601733258e-06,
+      "loss": 0.5358,
+      "step": 11039
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.849258668729484,
+      "learning_rate": 2.1366028676633534e-06,
+      "loss": 0.5159,
+      "step": 11040
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7024594752468707,
+      "learning_rate": 2.1357572554791166e-06,
+      "loss": 0.4733,
+      "step": 11041
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.602915841238694,
+      "learning_rate": 2.1349117652165384e-06,
+      "loss": 0.4291,
+      "step": 11042
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6994290517955272,
+      "learning_rate": 2.1340663969116117e-06,
+      "loss": 0.4836,
+      "step": 11043
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7290551463236109,
+      "learning_rate": 2.1332211506003215e-06,
+      "loss": 0.5759,
+      "step": 11044
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7593836673540703,
+      "learning_rate": 2.1323760263186442e-06,
+      "loss": 0.5106,
+      "step": 11045
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.569954565746356,
+      "learning_rate": 2.1315310241025576e-06,
+      "loss": 0.481,
+      "step": 11046
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7105864612712174,
+      "learning_rate": 2.130686143988027e-06,
+      "loss": 0.4888,
+      "step": 11047
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7373290698642639,
+      "learning_rate": 2.1298413860110207e-06,
+      "loss": 0.4938,
+      "step": 11048
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6343770581486133,
+      "learning_rate": 2.1289967502074927e-06,
+      "loss": 0.4557,
+      "step": 11049
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5774901260631604,
+      "learning_rate": 2.1281522366134007e-06,
+      "loss": 0.4714,
+      "step": 11050
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6454348805802032,
+      "learning_rate": 2.127307845264689e-06,
+      "loss": 0.5188,
+      "step": 11051
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.109335767711921,
+      "learning_rate": 2.126463576197304e-06,
+      "loss": 0.4535,
+      "step": 11052
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6783969564126169,
+      "learning_rate": 2.1256194294471806e-06,
+      "loss": 0.4939,
+      "step": 11053
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.766607386119204,
+      "learning_rate": 2.1247754050502545e-06,
+      "loss": 0.5071,
+      "step": 11054
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6764714887012292,
+      "learning_rate": 2.123931503042448e-06,
+      "loss": 0.4915,
+      "step": 11055
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.67241679643197,
+      "learning_rate": 2.1230877234596904e-06,
+      "loss": 0.4741,
+      "step": 11056
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6111770577975992,
+      "learning_rate": 2.122244066337893e-06,
+      "loss": 0.437,
+      "step": 11057
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6379618823997341,
+      "learning_rate": 2.1214005317129714e-06,
+      "loss": 0.4311,
+      "step": 11058
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6652716344344426,
+      "learning_rate": 2.120557119620829e-06,
+      "loss": 0.5059,
+      "step": 11059
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6668437273619943,
+      "learning_rate": 2.119713830097371e-06,
+      "loss": 0.5124,
+      "step": 11060
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7391629689550185,
+      "learning_rate": 2.118870663178488e-06,
+      "loss": 0.5147,
+      "step": 11061
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7573888819210415,
+      "learning_rate": 2.118027618900077e-06,
+      "loss": 0.557,
+      "step": 11062
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5660830013577599,
+      "learning_rate": 2.1171846972980185e-06,
+      "loss": 0.455,
+      "step": 11063
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5414229853801957,
+      "learning_rate": 2.116341898408198e-06,
+      "loss": 0.4536,
+      "step": 11064
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.656530204708304,
+      "learning_rate": 2.1154992222664857e-06,
+      "loss": 0.4856,
+      "step": 11065
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5606991332043892,
+      "learning_rate": 2.114656668908755e-06,
+      "loss": 0.4488,
+      "step": 11066
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6274954726238137,
+      "learning_rate": 2.113814238370867e-06,
+      "loss": 0.4774,
+      "step": 11067
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7467665757531308,
+      "learning_rate": 2.112971930688687e-06,
+      "loss": 0.5125,
+      "step": 11068
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6857575984261297,
+      "learning_rate": 2.112129745898065e-06,
+      "loss": 0.4938,
+      "step": 11069
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8540164108230414,
+      "learning_rate": 2.1112876840348527e-06,
+      "loss": 0.5532,
+      "step": 11070
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7300724933284729,
+      "learning_rate": 2.110445745134891e-06,
+      "loss": 0.5386,
+      "step": 11071
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6618732898394578,
+      "learning_rate": 2.109603929234022e-06,
+      "loss": 0.4534,
+      "step": 11072
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7967401665876949,
+      "learning_rate": 2.108762236368076e-06,
+      "loss": 0.4595,
+      "step": 11073
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6252598075712752,
+      "learning_rate": 2.1079206665728835e-06,
+      "loss": 0.4535,
+      "step": 11074
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.658759871928441,
+      "learning_rate": 2.1070792198842654e-06,
+      "loss": 0.4671,
+      "step": 11075
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7109059281672796,
+      "learning_rate": 2.106237896338042e-06,
+      "loss": 0.5544,
+      "step": 11076
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.825859182314527,
+      "learning_rate": 2.1053966959700223e-06,
+      "loss": 0.5077,
+      "step": 11077
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6402100340838714,
+      "learning_rate": 2.104555618816018e-06,
+      "loss": 0.4498,
+      "step": 11078
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7458329944559269,
+      "learning_rate": 2.1037146649118256e-06,
+      "loss": 0.5309,
+      "step": 11079
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7952858746286077,
+      "learning_rate": 2.1028738342932453e-06,
+      "loss": 0.5535,
+      "step": 11080
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8016345419733305,
+      "learning_rate": 2.1020331269960676e-06,
+      "loss": 0.4869,
+      "step": 11081
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7268533641558307,
+      "learning_rate": 2.101192543056081e-06,
+      "loss": 0.492,
+      "step": 11082
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7324600575608072,
+      "learning_rate": 2.100352082509063e-06,
+      "loss": 0.5313,
+      "step": 11083
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8198207889099383,
+      "learning_rate": 2.0995117453907927e-06,
+      "loss": 0.5854,
+      "step": 11084
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.74181732426242,
+      "learning_rate": 2.098671531737037e-06,
+      "loss": 0.5267,
+      "step": 11085
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8123453098551288,
+      "learning_rate": 2.0978314415835643e-06,
+      "loss": 0.5246,
+      "step": 11086
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.663872066440253,
+      "learning_rate": 2.096991474966132e-06,
+      "loss": 0.52,
+      "step": 11087
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5752085595551722,
+      "learning_rate": 2.0961516319204973e-06,
+      "loss": 0.4402,
+      "step": 11088
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5324093230432396,
+      "learning_rate": 2.095311912482406e-06,
+      "loss": 0.4791,
+      "step": 11089
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7119248843311933,
+      "learning_rate": 2.094472316687607e-06,
+      "loss": 0.4677,
+      "step": 11090
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7149301728187657,
+      "learning_rate": 2.0936328445718348e-06,
+      "loss": 0.5514,
+      "step": 11091
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7072166361555027,
+      "learning_rate": 2.0927934961708248e-06,
+      "loss": 0.5195,
+      "step": 11092
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6971844451125663,
+      "learning_rate": 2.0919542715203072e-06,
+      "loss": 0.5274,
+      "step": 11093
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5999122124993302,
+      "learning_rate": 2.0911151706560024e-06,
+      "loss": 0.4643,
+      "step": 11094
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5763592098279886,
+      "learning_rate": 2.0902761936136305e-06,
+      "loss": 0.4263,
+      "step": 11095
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5607236319769412,
+      "learning_rate": 2.0894373404289013e-06,
+      "loss": 0.445,
+      "step": 11096
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5850345854981462,
+      "learning_rate": 2.0885986111375234e-06,
+      "loss": 0.4734,
+      "step": 11097
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8683882082768954,
+      "learning_rate": 2.0877600057752014e-06,
+      "loss": 0.5556,
+      "step": 11098
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6896501195309843,
+      "learning_rate": 2.0869215243776276e-06,
+      "loss": 0.5399,
+      "step": 11099
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7537990629141019,
+      "learning_rate": 2.0860831669804978e-06,
+      "loss": 0.5455,
+      "step": 11100
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7181973726146353,
+      "learning_rate": 2.0852449336194943e-06,
+      "loss": 0.5286,
+      "step": 11101
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7061407701195933,
+      "learning_rate": 2.0844068243303006e-06,
+      "loss": 0.4676,
+      "step": 11102
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.657599665368253,
+      "learning_rate": 2.0835688391485903e-06,
+      "loss": 0.4559,
+      "step": 11103
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.738946946170603,
+      "learning_rate": 2.0827309781100354e-06,
+      "loss": 0.4881,
+      "step": 11104
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5833938737411153,
+      "learning_rate": 2.081893241250302e-06,
+      "loss": 0.4426,
+      "step": 11105
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6785732203369954,
+      "learning_rate": 2.081055628605046e-06,
+      "loss": 0.4466,
+      "step": 11106
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6813897254589607,
+      "learning_rate": 2.080218140209927e-06,
+      "loss": 0.4964,
+      "step": 11107
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6696063341332344,
+      "learning_rate": 2.079380776100589e-06,
+      "loss": 0.4914,
+      "step": 11108
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8094250466191285,
+      "learning_rate": 2.0785435363126806e-06,
+      "loss": 0.5014,
+      "step": 11109
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6141977248208548,
+      "learning_rate": 2.077706420881836e-06,
+      "loss": 0.5114,
+      "step": 11110
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7190070221409822,
+      "learning_rate": 2.076869429843693e-06,
+      "loss": 0.4776,
+      "step": 11111
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6745433180701251,
+      "learning_rate": 2.0760325632338753e-06,
+      "loss": 0.528,
+      "step": 11112
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7104035344480847,
+      "learning_rate": 2.0751958210880097e-06,
+      "loss": 0.5019,
+      "step": 11113
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.602695777505395,
+      "learning_rate": 2.07435920344171e-06,
+      "loss": 0.4308,
+      "step": 11114
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5926985940921856,
+      "learning_rate": 2.073522710330591e-06,
+      "loss": 0.5058,
+      "step": 11115
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7151406497260272,
+      "learning_rate": 2.0726863417902576e-06,
+      "loss": 0.5797,
+      "step": 11116
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.712438332720332,
+      "learning_rate": 2.0718500978563115e-06,
+      "loss": 0.4697,
+      "step": 11117
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5642295050730727,
+      "learning_rate": 2.0710139785643508e-06,
+      "loss": 0.4312,
+      "step": 11118
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6154997107422874,
+      "learning_rate": 2.0701779839499665e-06,
+      "loss": 0.4491,
+      "step": 11119
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8665712240284359,
+      "learning_rate": 2.069342114048741e-06,
+      "loss": 0.5318,
+      "step": 11120
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.664008458206658,
+      "learning_rate": 2.0685063688962587e-06,
+      "loss": 0.5069,
+      "step": 11121
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7600195362182983,
+      "learning_rate": 2.067670748528091e-06,
+      "loss": 0.4687,
+      "step": 11122
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6821995629661899,
+      "learning_rate": 2.0668352529798107e-06,
+      "loss": 0.4949,
+      "step": 11123
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6727349306651959,
+      "learning_rate": 2.0659998822869786e-06,
+      "loss": 0.4401,
+      "step": 11124
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6592809234476772,
+      "learning_rate": 2.065164636485158e-06,
+      "loss": 0.4829,
+      "step": 11125
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7127709604367063,
+      "learning_rate": 2.0643295156098988e-06,
+      "loss": 0.4981,
+      "step": 11126
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6462875817956245,
+      "learning_rate": 2.063494519696752e-06,
+      "loss": 0.4907,
+      "step": 11127
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7877815350100381,
+      "learning_rate": 2.062659648781259e-06,
+      "loss": 0.5668,
+      "step": 11128
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7131156581817333,
+      "learning_rate": 2.061824902898958e-06,
+      "loss": 0.4769,
+      "step": 11129
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5986380935803617,
+      "learning_rate": 2.0609902820853813e-06,
+      "loss": 0.4739,
+      "step": 11130
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7528864095873341,
+      "learning_rate": 2.060155786376059e-06,
+      "loss": 0.512,
+      "step": 11131
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6173057949871051,
+      "learning_rate": 2.0593214158065086e-06,
+      "loss": 0.4628,
+      "step": 11132
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6378019275944758,
+      "learning_rate": 2.0584871704122494e-06,
+      "loss": 0.4462,
+      "step": 11133
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7292249256808478,
+      "learning_rate": 2.0576530502287908e-06,
+      "loss": 0.4834,
+      "step": 11134
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.707451097709552,
+      "learning_rate": 2.056819055291641e-06,
+      "loss": 0.4663,
+      "step": 11135
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5947493844942456,
+      "learning_rate": 2.0559851856362973e-06,
+      "loss": 0.4755,
+      "step": 11136
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7038553669745728,
+      "learning_rate": 2.0551514412982583e-06,
+      "loss": 0.4593,
+      "step": 11137
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6810359109290924,
+      "learning_rate": 2.05431782231301e-06,
+      "loss": 0.4577,
+      "step": 11138
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8146909236585748,
+      "learning_rate": 2.0534843287160404e-06,
+      "loss": 0.4712,
+      "step": 11139
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7802695331532603,
+      "learning_rate": 2.0526509605428264e-06,
+      "loss": 0.5225,
+      "step": 11140
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7135752475066814,
+      "learning_rate": 2.0518177178288435e-06,
+      "loss": 0.4605,
+      "step": 11141
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5737366886981783,
+      "learning_rate": 2.0509846006095562e-06,
+      "loss": 0.4439,
+      "step": 11142
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5678085890581392,
+      "learning_rate": 2.0501516089204344e-06,
+      "loss": 0.445,
+      "step": 11143
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7178302723252884,
+      "learning_rate": 2.04931874279693e-06,
+      "loss": 0.5514,
+      "step": 11144
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7097429745694298,
+      "learning_rate": 2.0484860022745003e-06,
+      "loss": 0.4709,
+      "step": 11145
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6620569937354427,
+      "learning_rate": 2.0476533873885873e-06,
+      "loss": 0.4507,
+      "step": 11146
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6340940340417469,
+      "learning_rate": 2.046820898174637e-06,
+      "loss": 0.4386,
+      "step": 11147
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6325414899016335,
+      "learning_rate": 2.0459885346680825e-06,
+      "loss": 0.5396,
+      "step": 11148
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7611898111676605,
+      "learning_rate": 2.0451562969043582e-06,
+      "loss": 0.5393,
+      "step": 11149
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7448090154732241,
+      "learning_rate": 2.044324184918886e-06,
+      "loss": 0.5038,
+      "step": 11150
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6488845329716524,
+      "learning_rate": 2.04349219874709e-06,
+      "loss": 0.4581,
+      "step": 11151
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6951950446896673,
+      "learning_rate": 2.042660338424382e-06,
+      "loss": 0.4794,
+      "step": 11152
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.1265786487627243,
+      "learning_rate": 2.0418286039861752e-06,
+      "loss": 0.5148,
+      "step": 11153
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7184494276937451,
+      "learning_rate": 2.0409969954678695e-06,
+      "loss": 0.4835,
+      "step": 11154
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6291749993280014,
+      "learning_rate": 2.040165512904866e-06,
+      "loss": 0.4895,
+      "step": 11155
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5923525579805895,
+      "learning_rate": 2.0393341563325603e-06,
+      "loss": 0.421,
+      "step": 11156
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6847813843617118,
+      "learning_rate": 2.038502925786337e-06,
+      "loss": 0.464,
+      "step": 11157
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6029038851321042,
+      "learning_rate": 2.0376718213015824e-06,
+      "loss": 0.4372,
+      "step": 11158
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7022769911202208,
+      "learning_rate": 2.0368408429136697e-06,
+      "loss": 0.5035,
+      "step": 11159
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6747759573489848,
+      "learning_rate": 2.036009990657974e-06,
+      "loss": 0.4732,
+      "step": 11160
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7743521332684044,
+      "learning_rate": 2.035179264569863e-06,
+      "loss": 0.4949,
+      "step": 11161
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7944481008965669,
+      "learning_rate": 2.0343486646846945e-06,
+      "loss": 0.5759,
+      "step": 11162
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7901209155580213,
+      "learning_rate": 2.0335181910378286e-06,
+      "loss": 0.5553,
+      "step": 11163
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6756152331872225,
+      "learning_rate": 2.0326878436646118e-06,
+      "loss": 0.4994,
+      "step": 11164
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7415721565195527,
+      "learning_rate": 2.0318576226003932e-06,
+      "loss": 0.5608,
+      "step": 11165
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7322104642579267,
+      "learning_rate": 2.031027527880509e-06,
+      "loss": 0.4858,
+      "step": 11166
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7065383463815319,
+      "learning_rate": 2.0301975595402954e-06,
+      "loss": 0.4636,
+      "step": 11167
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5976634669660313,
+      "learning_rate": 2.029367717615084e-06,
+      "loss": 0.4799,
+      "step": 11168
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8025401793153247,
+      "learning_rate": 2.0285380021401934e-06,
+      "loss": 0.5876,
+      "step": 11169
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.671937427847929,
+      "learning_rate": 2.0277084131509468e-06,
+      "loss": 0.4363,
+      "step": 11170
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7865110059584666,
+      "learning_rate": 2.026878950682653e-06,
+      "loss": 0.5837,
+      "step": 11171
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8299314315854112,
+      "learning_rate": 2.0260496147706233e-06,
+      "loss": 0.5397,
+      "step": 11172
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7134080059598646,
+      "learning_rate": 2.025220405450157e-06,
+      "loss": 0.4863,
+      "step": 11173
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6397844984212663,
+      "learning_rate": 2.0243913227565526e-06,
+      "loss": 0.5193,
+      "step": 11174
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9453151750863827,
+      "learning_rate": 2.0235623667250996e-06,
+      "loss": 0.5469,
+      "step": 11175
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7859150893954437,
+      "learning_rate": 2.022733537391088e-06,
+      "loss": 0.5461,
+      "step": 11176
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.827419865591755,
+      "learning_rate": 2.021904834789793e-06,
+      "loss": 0.546,
+      "step": 11177
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7229751042227601,
+      "learning_rate": 2.0210762589564942e-06,
+      "loss": 0.5172,
+      "step": 11178
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6197193337111703,
+      "learning_rate": 2.0202478099264565e-06,
+      "loss": 0.4905,
+      "step": 11179
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7552588795551329,
+      "learning_rate": 2.0194194877349516e-06,
+      "loss": 0.4981,
+      "step": 11180
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8029246219288291,
+      "learning_rate": 2.0185912924172325e-06,
+      "loss": 0.5319,
+      "step": 11181
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6915913682263175,
+      "learning_rate": 2.017763224008556e-06,
+      "loss": 0.5061,
+      "step": 11182
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7326998335946943,
+      "learning_rate": 2.016935282544168e-06,
+      "loss": 0.5124,
+      "step": 11183
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7469400014112825,
+      "learning_rate": 2.0161074680593144e-06,
+      "loss": 0.5211,
+      "step": 11184
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6264426768938137,
+      "learning_rate": 2.0152797805892286e-06,
+      "loss": 0.4671,
+      "step": 11185
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5935187660313015,
+      "learning_rate": 2.014452220169147e-06,
+      "loss": 0.4217,
+      "step": 11186
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.782048419085864,
+      "learning_rate": 2.013624786834292e-06,
+      "loss": 0.5533,
+      "step": 11187
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.856726064302872,
+      "learning_rate": 2.012797480619888e-06,
+      "loss": 0.5277,
+      "step": 11188
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.9184237425366315,
+      "learning_rate": 2.011970301561148e-06,
+      "loss": 0.4777,
+      "step": 11189
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6465520845948611,
+      "learning_rate": 2.011143249693286e-06,
+      "loss": 0.4747,
+      "step": 11190
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6003114960282584,
+      "learning_rate": 2.0103163250515007e-06,
+      "loss": 0.4776,
+      "step": 11191
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7659828507649711,
+      "learning_rate": 2.009489527671e-06,
+      "loss": 0.5272,
+      "step": 11192
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6893242718725018,
+      "learning_rate": 2.0086628575869716e-06,
+      "loss": 0.449,
+      "step": 11193
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6823328161925785,
+      "learning_rate": 2.007836314834608e-06,
+      "loss": 0.4649,
+      "step": 11194
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.630098347005655,
+      "learning_rate": 2.0070098994490893e-06,
+      "loss": 0.4504,
+      "step": 11195
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6010447452932413,
+      "learning_rate": 2.006183611465597e-06,
+      "loss": 0.4465,
+      "step": 11196
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6568021747081857,
+      "learning_rate": 2.0053574509192996e-06,
+      "loss": 0.4214,
+      "step": 11197
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6959581616117472,
+      "learning_rate": 2.004531417845368e-06,
+      "loss": 0.5096,
+      "step": 11198
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6687655739177488,
+      "learning_rate": 2.0037055122789594e-06,
+      "loss": 0.518,
+      "step": 11199
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8096255649297281,
+      "learning_rate": 2.0028797342552346e-06,
+      "loss": 0.4608,
+      "step": 11200
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5901434615746427,
+      "learning_rate": 2.00205408380934e-06,
+      "loss": 0.4439,
+      "step": 11201
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6712716973862201,
+      "learning_rate": 2.0012285609764255e-06,
+      "loss": 0.4618,
+      "step": 11202
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5695432627942246,
+      "learning_rate": 2.0004031657916254e-06,
+      "loss": 0.3809,
+      "step": 11203
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5874289189650648,
+      "learning_rate": 1.9995778982900783e-06,
+      "loss": 0.4209,
+      "step": 11204
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5770341747333945,
+      "learning_rate": 1.9987527585069117e-06,
+      "loss": 0.3952,
+      "step": 11205
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5818497800879188,
+      "learning_rate": 1.997927746477251e-06,
+      "loss": 0.4737,
+      "step": 11206
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7840189837384709,
+      "learning_rate": 1.9971028622362116e-06,
+      "loss": 0.5592,
+      "step": 11207
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.652611209058209,
+      "learning_rate": 1.996278105818909e-06,
+      "loss": 0.4959,
+      "step": 11208
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7051516265529143,
+      "learning_rate": 1.9954534772604473e-06,
+      "loss": 0.5215,
+      "step": 11209
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7113256398763603,
+      "learning_rate": 1.994628976595932e-06,
+      "loss": 0.565,
+      "step": 11210
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6076305234717412,
+      "learning_rate": 1.993804603860455e-06,
+      "loss": 0.4279,
+      "step": 11211
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5894178483904796,
+      "learning_rate": 1.9929803590891115e-06,
+      "loss": 0.4303,
+      "step": 11212
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6180690102402564,
+      "learning_rate": 1.992156242316984e-06,
+      "loss": 0.4802,
+      "step": 11213
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8086175978414927,
+      "learning_rate": 1.991332253579154e-06,
+      "loss": 0.5127,
+      "step": 11214
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6086401440154163,
+      "learning_rate": 1.990508392910695e-06,
+      "loss": 0.4643,
+      "step": 11215
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7020827946931136,
+      "learning_rate": 1.9896846603466762e-06,
+      "loss": 0.4307,
+      "step": 11216
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6059436710261036,
+      "learning_rate": 1.988861055922164e-06,
+      "loss": 0.5241,
+      "step": 11217
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7541097747004496,
+      "learning_rate": 1.9880375796722116e-06,
+      "loss": 0.4994,
+      "step": 11218
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.56993310040587,
+      "learning_rate": 1.9872142316318775e-06,
+      "loss": 0.4395,
+      "step": 11219
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7569533892047728,
+      "learning_rate": 1.986391011836204e-06,
+      "loss": 0.5588,
+      "step": 11220
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6036030514550667,
+      "learning_rate": 1.985567920320235e-06,
+      "loss": 0.4207,
+      "step": 11221
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.580993050558816,
+      "learning_rate": 1.9847449571190087e-06,
+      "loss": 0.4573,
+      "step": 11222
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7089009313971928,
+      "learning_rate": 1.983922122267552e-06,
+      "loss": 0.4194,
+      "step": 11223
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6714513655693373,
+      "learning_rate": 1.983099415800895e-06,
+      "loss": 0.4894,
+      "step": 11224
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6338727037309118,
+      "learning_rate": 1.982276837754053e-06,
+      "loss": 0.504,
+      "step": 11225
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.705206243507172,
+      "learning_rate": 1.9814543881620447e-06,
+      "loss": 0.4273,
+      "step": 11226
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.657490437945099,
+      "learning_rate": 1.980632067059875e-06,
+      "loss": 0.4972,
+      "step": 11227
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7166418058104774,
+      "learning_rate": 1.97980987448255e-06,
+      "loss": 0.5376,
+      "step": 11228
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8066864628986806,
+      "learning_rate": 1.9789878104650694e-06,
+      "loss": 0.5398,
+      "step": 11229
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6102186772099529,
+      "learning_rate": 1.978165875042422e-06,
+      "loss": 0.4401,
+      "step": 11230
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6344805523278496,
+      "learning_rate": 1.977344068249599e-06,
+      "loss": 0.4665,
+      "step": 11231
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5943139674369584,
+      "learning_rate": 1.976522390121578e-06,
+      "loss": 0.4631,
+      "step": 11232
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7749393310510425,
+      "learning_rate": 1.9757008406933385e-06,
+      "loss": 0.546,
+      "step": 11233
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8174546481772259,
+      "learning_rate": 1.974879419999849e-06,
+      "loss": 0.5262,
+      "step": 11234
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7244544058728543,
+      "learning_rate": 1.9740581280760767e-06,
+      "loss": 0.5439,
+      "step": 11235
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7589498656694413,
+      "learning_rate": 1.9732369649569794e-06,
+      "loss": 0.5254,
+      "step": 11236
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6228498479634623,
+      "learning_rate": 1.9724159306775136e-06,
+      "loss": 0.4858,
+      "step": 11237
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7168217106689247,
+      "learning_rate": 1.971595025272625e-06,
+      "loss": 0.4725,
+      "step": 11238
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7366200843304248,
+      "learning_rate": 1.9707742487772607e-06,
+      "loss": 0.4036,
+      "step": 11239
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6148224541790992,
+      "learning_rate": 1.9699536012263546e-06,
+      "loss": 0.472,
+      "step": 11240
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7975209229405948,
+      "learning_rate": 1.9691330826548415e-06,
+      "loss": 0.5936,
+      "step": 11241
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7679398132762768,
+      "learning_rate": 1.9683126930976477e-06,
+      "loss": 0.5078,
+      "step": 11242
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7457473173654783,
+      "learning_rate": 1.967492432589696e-06,
+      "loss": 0.5207,
+      "step": 11243
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8982439231148711,
+      "learning_rate": 1.9666723011659e-06,
+      "loss": 0.5527,
+      "step": 11244
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7989003897825304,
+      "learning_rate": 1.965852298861172e-06,
+      "loss": 0.5473,
+      "step": 11245
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7330452786755091,
+      "learning_rate": 1.965032425710414e-06,
+      "loss": 0.5118,
+      "step": 11246
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6442375600589755,
+      "learning_rate": 1.9642126817485297e-06,
+      "loss": 0.4427,
+      "step": 11247
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6781118787117099,
+      "learning_rate": 1.9633930670104084e-06,
+      "loss": 0.4561,
+      "step": 11248
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5674129129708307,
+      "learning_rate": 1.9625735815309425e-06,
+      "loss": 0.4501,
+      "step": 11249
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8035437835222323,
+      "learning_rate": 1.9617542253450116e-06,
+      "loss": 0.5097,
+      "step": 11250
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6751218774101819,
+      "learning_rate": 1.960934998487496e-06,
+      "loss": 0.5032,
+      "step": 11251
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7490261223356152,
+      "learning_rate": 1.960115900993264e-06,
+      "loss": 0.5509,
+      "step": 11252
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.2838419789777458,
+      "learning_rate": 1.9592969328971844e-06,
+      "loss": 0.5794,
+      "step": 11253
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7122731410537773,
+      "learning_rate": 1.9584780942341177e-06,
+      "loss": 0.4885,
+      "step": 11254
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6522230509906171,
+      "learning_rate": 1.95765938503892e-06,
+      "loss": 0.436,
+      "step": 11255
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7400595324739233,
+      "learning_rate": 1.95684080534644e-06,
+      "loss": 0.5197,
+      "step": 11256
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.648825901543588,
+      "learning_rate": 1.956022355191523e-06,
+      "loss": 0.4806,
+      "step": 11257
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7492382274165247,
+      "learning_rate": 1.955204034609006e-06,
+      "loss": 0.5588,
+      "step": 11258
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7775395558613977,
+      "learning_rate": 1.9543858436337254e-06,
+      "loss": 0.5107,
+      "step": 11259
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6228703595391435,
+      "learning_rate": 1.953567782300505e-06,
+      "loss": 0.4804,
+      "step": 11260
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8130732016887502,
+      "learning_rate": 1.9527498506441704e-06,
+      "loss": 0.5168,
+      "step": 11261
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6351959209709755,
+      "learning_rate": 1.9519320486995362e-06,
+      "loss": 0.4897,
+      "step": 11262
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7671735089353404,
+      "learning_rate": 1.951114376501416e-06,
+      "loss": 0.4768,
+      "step": 11263
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6259057856947268,
+      "learning_rate": 1.9502968340846118e-06,
+      "loss": 0.4697,
+      "step": 11264
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7246265908714671,
+      "learning_rate": 1.9494794214839276e-06,
+      "loss": 0.5399,
+      "step": 11265
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9907033676110238,
+      "learning_rate": 1.9486621387341533e-06,
+      "loss": 0.5314,
+      "step": 11266
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7606284064873436,
+      "learning_rate": 1.9478449858700843e-06,
+      "loss": 0.5115,
+      "step": 11267
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7989713122800921,
+      "learning_rate": 1.947027962926499e-06,
+      "loss": 0.4791,
+      "step": 11268
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5817381244530586,
+      "learning_rate": 1.94621106993818e-06,
+      "loss": 0.4393,
+      "step": 11269
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7096321112885071,
+      "learning_rate": 1.9453943069398944e-06,
+      "loss": 0.4987,
+      "step": 11270
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.876752355322076,
+      "learning_rate": 1.944577673966415e-06,
+      "loss": 0.531,
+      "step": 11271
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7602148269568045,
+      "learning_rate": 1.943761171052497e-06,
+      "loss": 0.5576,
+      "step": 11272
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8329920170953107,
+      "learning_rate": 1.9429447982329024e-06,
+      "loss": 0.5284,
+      "step": 11273
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.684407161842788,
+      "learning_rate": 1.9421285555423767e-06,
+      "loss": 0.491,
+      "step": 11274
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7134658758564494,
+      "learning_rate": 1.9413124430156687e-06,
+      "loss": 0.4671,
+      "step": 11275
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.72173377341069,
+      "learning_rate": 1.9404964606875144e-06,
+      "loss": 0.5359,
+      "step": 11276
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7418848216152063,
+      "learning_rate": 1.93968060859265e-06,
+      "loss": 0.5247,
+      "step": 11277
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6680440353861876,
+      "learning_rate": 1.9388648867658017e-06,
+      "loss": 0.4564,
+      "step": 11278
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7389863164262619,
+      "learning_rate": 1.938049295241693e-06,
+      "loss": 0.4955,
+      "step": 11279
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5829233732668375,
+      "learning_rate": 1.9372338340550427e-06,
+      "loss": 0.4627,
+      "step": 11280
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8067337924033198,
+      "learning_rate": 1.936418503240559e-06,
+      "loss": 0.5247,
+      "step": 11281
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6413165594994173,
+      "learning_rate": 1.9356033028329524e-06,
+      "loss": 0.4939,
+      "step": 11282
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7492170614952702,
+      "learning_rate": 1.934788232866918e-06,
+      "loss": 0.5465,
+      "step": 11283
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5525736106000592,
+      "learning_rate": 1.933973293377154e-06,
+      "loss": 0.4346,
+      "step": 11284
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7094188651103669,
+      "learning_rate": 1.9331584843983513e-06,
+      "loss": 0.477,
+      "step": 11285
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5852625217234446,
+      "learning_rate": 1.9323438059651895e-06,
+      "loss": 0.4597,
+      "step": 11286
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5883877451196234,
+      "learning_rate": 1.9315292581123514e-06,
+      "loss": 0.462,
+      "step": 11287
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6345828121336127,
+      "learning_rate": 1.930714840874505e-06,
+      "loss": 0.487,
+      "step": 11288
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7379326876688959,
+      "learning_rate": 1.9299005542863213e-06,
+      "loss": 0.5374,
+      "step": 11289
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5923668077726157,
+      "learning_rate": 1.9290863983824594e-06,
+      "loss": 0.4956,
+      "step": 11290
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.449338058742408,
+      "learning_rate": 1.9282723731975754e-06,
+      "loss": 0.507,
+      "step": 11291
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6693495554699556,
+      "learning_rate": 1.927458478766323e-06,
+      "loss": 0.4585,
+      "step": 11292
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9453083517928659,
+      "learning_rate": 1.9266447151233432e-06,
+      "loss": 0.53,
+      "step": 11293
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6765940905558698,
+      "learning_rate": 1.925831082303278e-06,
+      "loss": 0.5118,
+      "step": 11294
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7102414801322271,
+      "learning_rate": 1.925017580340759e-06,
+      "loss": 0.4205,
+      "step": 11295
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6814300141816841,
+      "learning_rate": 1.924204209270416e-06,
+      "loss": 0.4422,
+      "step": 11296
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6169935658259402,
+      "learning_rate": 1.9233909691268705e-06,
+      "loss": 0.5047,
+      "step": 11297
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6935651286225928,
+      "learning_rate": 1.922577859944741e-06,
+      "loss": 0.5163,
+      "step": 11298
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6450800111967042,
+      "learning_rate": 1.921764881758636e-06,
+      "loss": 0.4421,
+      "step": 11299
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.620304074890015,
+      "learning_rate": 1.9209520346031656e-06,
+      "loss": 0.4345,
+      "step": 11300
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7315790932606675,
+      "learning_rate": 1.9201393185129256e-06,
+      "loss": 0.4929,
+      "step": 11301
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7004323220145465,
+      "learning_rate": 1.919326733522515e-06,
+      "loss": 0.5108,
+      "step": 11302
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6271859644453363,
+      "learning_rate": 1.918514279666518e-06,
+      "loss": 0.451,
+      "step": 11303
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6497526757629732,
+      "learning_rate": 1.9177019569795235e-06,
+      "loss": 0.4695,
+      "step": 11304
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7336140160469972,
+      "learning_rate": 1.916889765496106e-06,
+      "loss": 0.534,
+      "step": 11305
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.662297858676538,
+      "learning_rate": 1.9160777052508404e-06,
+      "loss": 0.46,
+      "step": 11306
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6510159096884519,
+      "learning_rate": 1.915265776278291e-06,
+      "loss": 0.5155,
+      "step": 11307
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6823491584859847,
+      "learning_rate": 1.9144539786130212e-06,
+      "loss": 0.4816,
+      "step": 11308
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7294871088530409,
+      "learning_rate": 1.913642312289584e-06,
+      "loss": 0.5134,
+      "step": 11309
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5819265618427858,
+      "learning_rate": 1.912830777342533e-06,
+      "loss": 0.4401,
+      "step": 11310
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6078669650032759,
+      "learning_rate": 1.9120193738064088e-06,
+      "loss": 0.4367,
+      "step": 11311
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9069761228682602,
+      "learning_rate": 1.9112081017157535e-06,
+      "loss": 0.5096,
+      "step": 11312
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6029898006326574,
+      "learning_rate": 1.910396961105097e-06,
+      "loss": 0.4402,
+      "step": 11313
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6438412119564003,
+      "learning_rate": 1.909585952008971e-06,
+      "loss": 0.4172,
+      "step": 11314
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5404246754231471,
+      "learning_rate": 1.908775074461892e-06,
+      "loss": 0.4303,
+      "step": 11315
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5388010883116349,
+      "learning_rate": 1.9079643284983836e-06,
+      "loss": 0.3819,
+      "step": 11316
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.605711647623545,
+      "learning_rate": 1.9071537141529512e-06,
+      "loss": 0.4336,
+      "step": 11317
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6829792860415591,
+      "learning_rate": 1.9063432314601033e-06,
+      "loss": 0.5251,
+      "step": 11318
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5740065416349589,
+      "learning_rate": 1.9055328804543367e-06,
+      "loss": 0.4714,
+      "step": 11319
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.716931513077479,
+      "learning_rate": 1.904722661170148e-06,
+      "loss": 0.5264,
+      "step": 11320
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.671637368678671,
+      "learning_rate": 1.9039125736420233e-06,
+      "loss": 0.4914,
+      "step": 11321
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.620428701543499,
+      "learning_rate": 1.9031026179044488e-06,
+      "loss": 0.4891,
+      "step": 11322
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.768085628408398,
+      "learning_rate": 1.9022927939918972e-06,
+      "loss": 0.5227,
+      "step": 11323
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6760972529286847,
+      "learning_rate": 1.9014831019388447e-06,
+      "loss": 0.4718,
+      "step": 11324
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6478800558768326,
+      "learning_rate": 1.9006735417797528e-06,
+      "loss": 0.4785,
+      "step": 11325
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.647028749615993,
+      "learning_rate": 1.8998641135490863e-06,
+      "loss": 0.4857,
+      "step": 11326
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6887815457774414,
+      "learning_rate": 1.899054817281296e-06,
+      "loss": 0.4534,
+      "step": 11327
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6492984440097781,
+      "learning_rate": 1.898245653010833e-06,
+      "loss": 0.4748,
+      "step": 11328
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.585218737768615,
+      "learning_rate": 1.8974366207721407e-06,
+      "loss": 0.4185,
+      "step": 11329
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6147387354800559,
+      "learning_rate": 1.8966277205996592e-06,
+      "loss": 0.4824,
+      "step": 11330
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5514378080388994,
+      "learning_rate": 1.8958189525278164e-06,
+      "loss": 0.4462,
+      "step": 11331
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6715572472281072,
+      "learning_rate": 1.8950103165910433e-06,
+      "loss": 0.425,
+      "step": 11332
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6701754334618644,
+      "learning_rate": 1.8942018128237571e-06,
+      "loss": 0.4713,
+      "step": 11333
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7383832353037361,
+      "learning_rate": 1.8933934412603766e-06,
+      "loss": 0.493,
+      "step": 11334
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7137313548136808,
+      "learning_rate": 1.892585201935308e-06,
+      "loss": 0.4885,
+      "step": 11335
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.654454215507374,
+      "learning_rate": 1.89177709488296e-06,
+      "loss": 0.4481,
+      "step": 11336
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6013489032851089,
+      "learning_rate": 1.890969120137726e-06,
+      "loss": 0.4469,
+      "step": 11337
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6710988442836858,
+      "learning_rate": 1.890161277734004e-06,
+      "loss": 0.5161,
+      "step": 11338
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7589133859914434,
+      "learning_rate": 1.889353567706177e-06,
+      "loss": 0.4744,
+      "step": 11339
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5587001051874372,
+      "learning_rate": 1.8885459900886282e-06,
+      "loss": 0.4116,
+      "step": 11340
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6221389837172623,
+      "learning_rate": 1.8877385449157354e-06,
+      "loss": 0.4751,
+      "step": 11341
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7048839085214085,
+      "learning_rate": 1.8869312322218658e-06,
+      "loss": 0.4974,
+      "step": 11342
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6847705843990224,
+      "learning_rate": 1.8861240520413876e-06,
+      "loss": 0.5328,
+      "step": 11343
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6374377065226372,
+      "learning_rate": 1.8853170044086562e-06,
+      "loss": 0.4914,
+      "step": 11344
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8486749525142614,
+      "learning_rate": 1.8845100893580264e-06,
+      "loss": 0.524,
+      "step": 11345
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6157314279655526,
+      "learning_rate": 1.8837033069238487e-06,
+      "loss": 0.4885,
+      "step": 11346
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7449133394383595,
+      "learning_rate": 1.8828966571404606e-06,
+      "loss": 0.5735,
+      "step": 11347
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7700475046969686,
+      "learning_rate": 1.8820901400422032e-06,
+      "loss": 0.4702,
+      "step": 11348
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6070447399975536,
+      "learning_rate": 1.8812837556634033e-06,
+      "loss": 0.4492,
+      "step": 11349
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6149711047935372,
+      "learning_rate": 1.8804775040383894e-06,
+      "loss": 0.4682,
+      "step": 11350
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6203758570810054,
+      "learning_rate": 1.879671385201478e-06,
+      "loss": 0.4655,
+      "step": 11351
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5958757086161411,
+      "learning_rate": 1.878865399186985e-06,
+      "loss": 0.4392,
+      "step": 11352
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7035927179298433,
+      "learning_rate": 1.8780595460292195e-06,
+      "loss": 0.5019,
+      "step": 11353
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6023035148596293,
+      "learning_rate": 1.8772538257624812e-06,
+      "loss": 0.4745,
+      "step": 11354
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5122257746860032,
+      "learning_rate": 1.8764482384210703e-06,
+      "loss": 0.4242,
+      "step": 11355
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6360494899344628,
+      "learning_rate": 1.8756427840392749e-06,
+      "loss": 0.4676,
+      "step": 11356
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.787876292155766,
+      "learning_rate": 1.8748374626513838e-06,
+      "loss": 0.517,
+      "step": 11357
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7996470719930312,
+      "learning_rate": 1.8740322742916738e-06,
+      "loss": 0.4522,
+      "step": 11358
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7036255965940895,
+      "learning_rate": 1.8732272189944224e-06,
+      "loss": 0.5299,
+      "step": 11359
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.759481931586932,
+      "learning_rate": 1.8724222967938944e-06,
+      "loss": 0.5199,
+      "step": 11360
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.695578461214734,
+      "learning_rate": 1.871617507724357e-06,
+      "loss": 0.5408,
+      "step": 11361
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7121033471142931,
+      "learning_rate": 1.8708128518200635e-06,
+      "loss": 0.4968,
+      "step": 11362
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.695666574665368,
+      "learning_rate": 1.8700083291152692e-06,
+      "loss": 0.4235,
+      "step": 11363
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6810425887666096,
+      "learning_rate": 1.8692039396442169e-06,
+      "loss": 0.5027,
+      "step": 11364
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8222239671783307,
+      "learning_rate": 1.8683996834411483e-06,
+      "loss": 0.5252,
+      "step": 11365
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8050408408150342,
+      "learning_rate": 1.8675955605402979e-06,
+      "loss": 0.5058,
+      "step": 11366
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8433880324569696,
+      "learning_rate": 1.866791570975896e-06,
+      "loss": 0.5451,
+      "step": 11367
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8072636368134379,
+      "learning_rate": 1.8659877147821637e-06,
+      "loss": 0.5163,
+      "step": 11368
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6638410952524344,
+      "learning_rate": 1.8651839919933206e-06,
+      "loss": 0.544,
+      "step": 11369
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7098494588444912,
+      "learning_rate": 1.8643804026435758e-06,
+      "loss": 0.528,
+      "step": 11370
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5987182590874014,
+      "learning_rate": 1.8635769467671394e-06,
+      "loss": 0.4282,
+      "step": 11371
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6676643531043296,
+      "learning_rate": 1.862773624398208e-06,
+      "loss": 0.4383,
+      "step": 11372
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8879658039018834,
+      "learning_rate": 1.8619704355709795e-06,
+      "loss": 0.5059,
+      "step": 11373
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8757993116504977,
+      "learning_rate": 1.8611673803196407e-06,
+      "loss": 0.5329,
+      "step": 11374
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8558627979305645,
+      "learning_rate": 1.8603644586783775e-06,
+      "loss": 0.5098,
+      "step": 11375
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6379557817780248,
+      "learning_rate": 1.8595616706813652e-06,
+      "loss": 0.4782,
+      "step": 11376
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6029648904436841,
+      "learning_rate": 1.8587590163627773e-06,
+      "loss": 0.4839,
+      "step": 11377
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.805259381730763,
+      "learning_rate": 1.8579564957567791e-06,
+      "loss": 0.5867,
+      "step": 11378
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7387227870112327,
+      "learning_rate": 1.8571541088975347e-06,
+      "loss": 0.5359,
+      "step": 11379
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6835881395539167,
+      "learning_rate": 1.8563518558191957e-06,
+      "loss": 0.4796,
+      "step": 11380
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7993258912990056,
+      "learning_rate": 1.8555497365559133e-06,
+      "loss": 0.5667,
+      "step": 11381
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6453699321625503,
+      "learning_rate": 1.8547477511418288e-06,
+      "loss": 0.422,
+      "step": 11382
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6184484823844142,
+      "learning_rate": 1.8539458996110838e-06,
+      "loss": 0.4446,
+      "step": 11383
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7343103680563909,
+      "learning_rate": 1.8531441819978069e-06,
+      "loss": 0.5102,
+      "step": 11384
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5681145579544644,
+      "learning_rate": 1.8523425983361276e-06,
+      "loss": 0.438,
+      "step": 11385
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6409367061892022,
+      "learning_rate": 1.8515411486601637e-06,
+      "loss": 0.5028,
+      "step": 11386
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7225861684427305,
+      "learning_rate": 1.8507398330040343e-06,
+      "loss": 0.4309,
+      "step": 11387
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6597519877532663,
+      "learning_rate": 1.849938651401844e-06,
+      "loss": 0.4718,
+      "step": 11388
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6714276098028268,
+      "learning_rate": 1.849137603887702e-06,
+      "loss": 0.513,
+      "step": 11389
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5961933529361594,
+      "learning_rate": 1.848336690495699e-06,
+      "loss": 0.4208,
+      "step": 11390
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6045272297501535,
+      "learning_rate": 1.8475359112599362e-06,
+      "loss": 0.3906,
+      "step": 11391
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8572888042345367,
+      "learning_rate": 1.8467352662144934e-06,
+      "loss": 0.4952,
+      "step": 11392
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6472699547003887,
+      "learning_rate": 1.8459347553934559e-06,
+      "loss": 0.4775,
+      "step": 11393
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6696770115024709,
+      "learning_rate": 1.8451343788308952e-06,
+      "loss": 0.5135,
+      "step": 11394
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8168211517611075,
+      "learning_rate": 1.844334136560884e-06,
+      "loss": 0.5001,
+      "step": 11395
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6928336793428703,
+      "learning_rate": 1.843534028617483e-06,
+      "loss": 0.4166,
+      "step": 11396
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6208105925383309,
+      "learning_rate": 1.8427340550347533e-06,
+      "loss": 0.4772,
+      "step": 11397
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6133442560191401,
+      "learning_rate": 1.8419342158467441e-06,
+      "loss": 0.4637,
+      "step": 11398
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5850717184361476,
+      "learning_rate": 1.8411345110875057e-06,
+      "loss": 0.4364,
+      "step": 11399
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6289716505473774,
+      "learning_rate": 1.8403349407910742e-06,
+      "loss": 0.4291,
+      "step": 11400
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6269852948282604,
+      "learning_rate": 1.8395355049914897e-06,
+      "loss": 0.435,
+      "step": 11401
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5992514811941154,
+      "learning_rate": 1.8387362037227779e-06,
+      "loss": 0.4228,
+      "step": 11402
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6669948651458143,
+      "learning_rate": 1.8379370370189636e-06,
+      "loss": 0.4457,
+      "step": 11403
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5847273644145587,
+      "learning_rate": 1.8371380049140675e-06,
+      "loss": 0.458,
+      "step": 11404
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7427856648280324,
+      "learning_rate": 1.8363391074420972e-06,
+      "loss": 0.5212,
+      "step": 11405
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8804268564421519,
+      "learning_rate": 1.8355403446370612e-06,
+      "loss": 0.5632,
+      "step": 11406
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7269331058491464,
+      "learning_rate": 1.8347417165329628e-06,
+      "loss": 0.5342,
+      "step": 11407
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7447227093176003,
+      "learning_rate": 1.8339432231637928e-06,
+      "loss": 0.4994,
+      "step": 11408
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6879438327484899,
+      "learning_rate": 1.8331448645635441e-06,
+      "loss": 0.4818,
+      "step": 11409
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6820188400244761,
+      "learning_rate": 1.8323466407661972e-06,
+      "loss": 0.4584,
+      "step": 11410
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.790304865458631,
+      "learning_rate": 1.8315485518057335e-06,
+      "loss": 0.4753,
+      "step": 11411
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8620632386124969,
+      "learning_rate": 1.8307505977161206e-06,
+      "loss": 0.4327,
+      "step": 11412
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.814672743559231,
+      "learning_rate": 1.8299527785313292e-06,
+      "loss": 0.5779,
+      "step": 11413
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7305306394057026,
+      "learning_rate": 1.829155094285317e-06,
+      "loss": 0.5228,
+      "step": 11414
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6052248339321922,
+      "learning_rate": 1.8283575450120389e-06,
+      "loss": 0.4842,
+      "step": 11415
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7594348748008597,
+      "learning_rate": 1.827560130745447e-06,
+      "loss": 0.566,
+      "step": 11416
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7507693321932712,
+      "learning_rate": 1.8267628515194813e-06,
+      "loss": 0.5104,
+      "step": 11417
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7326304713664651,
+      "learning_rate": 1.8259657073680826e-06,
+      "loss": 0.5434,
+      "step": 11418
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6851379089020471,
+      "learning_rate": 1.8251686983251788e-06,
+      "loss": 0.5361,
+      "step": 11419
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7546095013146231,
+      "learning_rate": 1.8243718244247e-06,
+      "loss": 0.5257,
+      "step": 11420
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7441539723313164,
+      "learning_rate": 1.8235750857005629e-06,
+      "loss": 0.5138,
+      "step": 11421
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6026725448776842,
+      "learning_rate": 1.822778482186686e-06,
+      "loss": 0.4859,
+      "step": 11422
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.752596168333072,
+      "learning_rate": 1.8219820139169741e-06,
+      "loss": 0.5233,
+      "step": 11423
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6210477534583401,
+      "learning_rate": 1.8211856809253348e-06,
+      "loss": 0.4774,
+      "step": 11424
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7279576260642828,
+      "learning_rate": 1.820389483245661e-06,
+      "loss": 0.4624,
+      "step": 11425
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6632562080822728,
+      "learning_rate": 1.8195934209118477e-06,
+      "loss": 0.4616,
+      "step": 11426
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6595600738049575,
+      "learning_rate": 1.8187974939577763e-06,
+      "loss": 0.4518,
+      "step": 11427
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6129363296267166,
+      "learning_rate": 1.818001702417333e-06,
+      "loss": 0.4411,
+      "step": 11428
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5890918786735774,
+      "learning_rate": 1.8172060463243874e-06,
+      "loss": 0.445,
+      "step": 11429
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7662057066416829,
+      "learning_rate": 1.816410525712811e-06,
+      "loss": 0.5797,
+      "step": 11430
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7353500142376335,
+      "learning_rate": 1.8156151406164645e-06,
+      "loss": 0.5474,
+      "step": 11431
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7961389928726098,
+      "learning_rate": 1.8148198910692066e-06,
+      "loss": 0.5096,
+      "step": 11432
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8054147583020271,
+      "learning_rate": 1.8140247771048863e-06,
+      "loss": 0.5398,
+      "step": 11433
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6711575202310909,
+      "learning_rate": 1.8132297987573516e-06,
+      "loss": 0.4428,
+      "step": 11434
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5596974355278252,
+      "learning_rate": 1.8124349560604394e-06,
+      "loss": 0.4393,
+      "step": 11435
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6128956121929855,
+      "learning_rate": 1.8116402490479867e-06,
+      "loss": 0.427,
+      "step": 11436
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6819978671117611,
+      "learning_rate": 1.8108456777538185e-06,
+      "loss": 0.4981,
+      "step": 11437
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7662783071339638,
+      "learning_rate": 1.8100512422117606e-06,
+      "loss": 0.5706,
+      "step": 11438
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.685410470491554,
+      "learning_rate": 1.809256942455624e-06,
+      "loss": 0.5081,
+      "step": 11439
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6449733075945355,
+      "learning_rate": 1.8084627785192265e-06,
+      "loss": 0.4794,
+      "step": 11440
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7363273460430979,
+      "learning_rate": 1.8076687504363683e-06,
+      "loss": 0.5689,
+      "step": 11441
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8143814366111622,
+      "learning_rate": 1.8068748582408518e-06,
+      "loss": 0.4884,
+      "step": 11442
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.615905581501184,
+      "learning_rate": 1.806081101966467e-06,
+      "loss": 0.4912,
+      "step": 11443
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6070018823252646,
+      "learning_rate": 1.8052874816470056e-06,
+      "loss": 0.5057,
+      "step": 11444
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7045085296061119,
+      "learning_rate": 1.804493997316245e-06,
+      "loss": 0.5331,
+      "step": 11445
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6514146328919574,
+      "learning_rate": 1.8037006490079656e-06,
+      "loss": 0.4335,
+      "step": 11446
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7368975197273516,
+      "learning_rate": 1.8029074367559335e-06,
+      "loss": 0.4163,
+      "step": 11447
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6413597586439923,
+      "learning_rate": 1.8021143605939173e-06,
+      "loss": 0.4392,
+      "step": 11448
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6562755475241322,
+      "learning_rate": 1.8013214205556722e-06,
+      "loss": 0.4934,
+      "step": 11449
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5744958712650743,
+      "learning_rate": 1.8005286166749548e-06,
+      "loss": 0.4319,
+      "step": 11450
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5898944792749423,
+      "learning_rate": 1.7997359489855083e-06,
+      "loss": 0.4334,
+      "step": 11451
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5474595776835868,
+      "learning_rate": 1.798943417521075e-06,
+      "loss": 0.4424,
+      "step": 11452
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7212489705428717,
+      "learning_rate": 1.7981510223153919e-06,
+      "loss": 0.4679,
+      "step": 11453
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6659344670590249,
+      "learning_rate": 1.79735876340219e-06,
+      "loss": 0.4797,
+      "step": 11454
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5789524318917341,
+      "learning_rate": 1.7965666408151894e-06,
+      "loss": 0.4792,
+      "step": 11455
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7589410901965923,
+      "learning_rate": 1.7957746545881121e-06,
+      "loss": 0.5447,
+      "step": 11456
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.0046614548571045,
+      "learning_rate": 1.7949828047546668e-06,
+      "loss": 0.5296,
+      "step": 11457
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7933823037260812,
+      "learning_rate": 1.7941910913485633e-06,
+      "loss": 0.5279,
+      "step": 11458
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7876437667701834,
+      "learning_rate": 1.7933995144034994e-06,
+      "loss": 0.4971,
+      "step": 11459
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5719763621423096,
+      "learning_rate": 1.7926080739531727e-06,
+      "loss": 0.4383,
+      "step": 11460
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6202015258602679,
+      "learning_rate": 1.7918167700312695e-06,
+      "loss": 0.4718,
+      "step": 11461
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5853265472101226,
+      "learning_rate": 1.7910256026714756e-06,
+      "loss": 0.5024,
+      "step": 11462
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.797943078973722,
+      "learning_rate": 1.790234571907466e-06,
+      "loss": 0.5613,
+      "step": 11463
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.831896870677325,
+      "learning_rate": 1.789443677772914e-06,
+      "loss": 0.5199,
+      "step": 11464
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.607534823891539,
+      "learning_rate": 1.7886529203014864e-06,
+      "loss": 0.4062,
+      "step": 11465
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6825610866584609,
+      "learning_rate": 1.7878622995268407e-06,
+      "loss": 0.5457,
+      "step": 11466
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7460391223936172,
+      "learning_rate": 1.7870718154826338e-06,
+      "loss": 0.54,
+      "step": 11467
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8683466754217286,
+      "learning_rate": 1.7862814682025115e-06,
+      "loss": 0.4967,
+      "step": 11468
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8502756211600191,
+      "learning_rate": 1.7854912577201167e-06,
+      "loss": 0.5297,
+      "step": 11469
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7780664606821395,
+      "learning_rate": 1.7847011840690892e-06,
+      "loss": 0.5598,
+      "step": 11470
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7337613261122693,
+      "learning_rate": 1.7839112472830562e-06,
+      "loss": 0.4994,
+      "step": 11471
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8168542737263976,
+      "learning_rate": 1.7831214473956454e-06,
+      "loss": 0.5212,
+      "step": 11472
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8352738925400326,
+      "learning_rate": 1.782331784440473e-06,
+      "loss": 0.5519,
+      "step": 11473
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8352580204131053,
+      "learning_rate": 1.7815422584511566e-06,
+      "loss": 0.5121,
+      "step": 11474
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7001900334173825,
+      "learning_rate": 1.7807528694612996e-06,
+      "loss": 0.4768,
+      "step": 11475
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8445488655757379,
+      "learning_rate": 1.779963617504506e-06,
+      "loss": 0.5117,
+      "step": 11476
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7406429898031691,
+      "learning_rate": 1.779174502614373e-06,
+      "loss": 0.498,
+      "step": 11477
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7893810111282417,
+      "learning_rate": 1.7783855248244875e-06,
+      "loss": 0.5518,
+      "step": 11478
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7855546277686893,
+      "learning_rate": 1.7775966841684367e-06,
+      "loss": 0.4456,
+      "step": 11479
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6855011200289312,
+      "learning_rate": 1.7768079806797967e-06,
+      "loss": 0.4914,
+      "step": 11480
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6938643355402148,
+      "learning_rate": 1.7760194143921423e-06,
+      "loss": 0.4794,
+      "step": 11481
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6574024281054431,
+      "learning_rate": 1.7752309853390377e-06,
+      "loss": 0.428,
+      "step": 11482
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5673485561577188,
+      "learning_rate": 1.7744426935540464e-06,
+      "loss": 0.4476,
+      "step": 11483
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6271509276827958,
+      "learning_rate": 1.7736545390707205e-06,
+      "loss": 0.5151,
+      "step": 11484
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6918206223414506,
+      "learning_rate": 1.7728665219226133e-06,
+      "loss": 0.5171,
+      "step": 11485
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6263608028548525,
+      "learning_rate": 1.7720786421432635e-06,
+      "loss": 0.5046,
+      "step": 11486
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6465377676486046,
+      "learning_rate": 1.771290899766213e-06,
+      "loss": 0.4291,
+      "step": 11487
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6907595096503566,
+      "learning_rate": 1.7705032948249895e-06,
+      "loss": 0.4566,
+      "step": 11488
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7075940045692216,
+      "learning_rate": 1.7697158273531207e-06,
+      "loss": 0.5063,
+      "step": 11489
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6813332020786361,
+      "learning_rate": 1.768928497384127e-06,
+      "loss": 0.5052,
+      "step": 11490
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7446828431298997,
+      "learning_rate": 1.7681413049515233e-06,
+      "loss": 0.4687,
+      "step": 11491
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5743777001582568,
+      "learning_rate": 1.7673542500888157e-06,
+      "loss": 0.4542,
+      "step": 11492
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.797217248313285,
+      "learning_rate": 1.7665673328295085e-06,
+      "loss": 0.5179,
+      "step": 11493
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.760027149228795,
+      "learning_rate": 1.7657805532070966e-06,
+      "loss": 0.485,
+      "step": 11494
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7939131495353275,
+      "learning_rate": 1.7649939112550724e-06,
+      "loss": 0.5341,
+      "step": 11495
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7116622065175608,
+      "learning_rate": 1.7642074070069182e-06,
+      "loss": 0.4707,
+      "step": 11496
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5997526212237774,
+      "learning_rate": 1.7634210404961165e-06,
+      "loss": 0.4579,
+      "step": 11497
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6592923954663346,
+      "learning_rate": 1.762634811756137e-06,
+      "loss": 0.4747,
+      "step": 11498
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5720989469457686,
+      "learning_rate": 1.7618487208204498e-06,
+      "loss": 0.4382,
+      "step": 11499
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5637432305131495,
+      "learning_rate": 1.7610627677225129e-06,
+      "loss": 0.4027,
+      "step": 11500
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6003834310857484,
+      "learning_rate": 1.7602769524957842e-06,
+      "loss": 0.4462,
+      "step": 11501
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.678516916390727,
+      "learning_rate": 1.7594912751737132e-06,
+      "loss": 0.4399,
+      "step": 11502
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.609378781142807,
+      "learning_rate": 1.7587057357897447e-06,
+      "loss": 0.44,
+      "step": 11503
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7350987040078207,
+      "learning_rate": 1.7579203343773138e-06,
+      "loss": 0.4426,
+      "step": 11504
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.690981948533854,
+      "learning_rate": 1.757135070969856e-06,
+      "loss": 0.494,
+      "step": 11505
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6602228679499301,
+      "learning_rate": 1.756349945600793e-06,
+      "loss": 0.464,
+      "step": 11506
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6912428253486794,
+      "learning_rate": 1.75556495830355e-06,
+      "loss": 0.4572,
+      "step": 11507
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6757033083216948,
+      "learning_rate": 1.7547801091115374e-06,
+      "loss": 0.4481,
+      "step": 11508
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6004194983371172,
+      "learning_rate": 1.7539953980581665e-06,
+      "loss": 0.411,
+      "step": 11509
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6629981984011224,
+      "learning_rate": 1.753210825176837e-06,
+      "loss": 0.4508,
+      "step": 11510
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5526800456809966,
+      "learning_rate": 1.7524263905009499e-06,
+      "loss": 0.4468,
+      "step": 11511
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5990447698490425,
+      "learning_rate": 1.7516420940638912e-06,
+      "loss": 0.4835,
+      "step": 11512
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6169888985794139,
+      "learning_rate": 1.7508579358990501e-06,
+      "loss": 0.4477,
+      "step": 11513
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5892362134564003,
+      "learning_rate": 1.7500739160398006e-06,
+      "loss": 0.4263,
+      "step": 11514
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5534155623237008,
+      "learning_rate": 1.749290034519523e-06,
+      "loss": 0.4296,
+      "step": 11515
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.701035013101386,
+      "learning_rate": 1.7485062913715784e-06,
+      "loss": 0.5216,
+      "step": 11516
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7680250293511391,
+      "learning_rate": 1.747722686629333e-06,
+      "loss": 0.5052,
+      "step": 11517
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7417850421576466,
+      "learning_rate": 1.746939220326138e-06,
+      "loss": 0.555,
+      "step": 11518
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7258841808079408,
+      "learning_rate": 1.746155892495347e-06,
+      "loss": 0.5377,
+      "step": 11519
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7553149431842945,
+      "learning_rate": 1.7453727031703e-06,
+      "loss": 0.5544,
+      "step": 11520
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8202925738083765,
+      "learning_rate": 1.7445896523843386e-06,
+      "loss": 0.5436,
+      "step": 11521
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7932840073403957,
+      "learning_rate": 1.7438067401707909e-06,
+      "loss": 0.5222,
+      "step": 11522
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6139123674514075,
+      "learning_rate": 1.7430239665629866e-06,
+      "loss": 0.5099,
+      "step": 11523
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.3698117525881095,
+      "learning_rate": 1.7422413315942433e-06,
+      "loss": 0.5158,
+      "step": 11524
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.604257915599999,
+      "learning_rate": 1.7414588352978774e-06,
+      "loss": 0.4342,
+      "step": 11525
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7645491208958822,
+      "learning_rate": 1.7406764777071938e-06,
+      "loss": 0.4932,
+      "step": 11526
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6857563844302612,
+      "learning_rate": 1.739894258855498e-06,
+      "loss": 0.5499,
+      "step": 11527
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5865634869951317,
+      "learning_rate": 1.7391121787760873e-06,
+      "loss": 0.4653,
+      "step": 11528
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6614811942942823,
+      "learning_rate": 1.7383302375022493e-06,
+      "loss": 0.5516,
+      "step": 11529
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7539716103220729,
+      "learning_rate": 1.73754843506727e-06,
+      "loss": 0.4831,
+      "step": 11530
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5865375857353755,
+      "learning_rate": 1.73676677150443e-06,
+      "loss": 0.4301,
+      "step": 11531
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6760179459749036,
+      "learning_rate": 1.7359852468469995e-06,
+      "loss": 0.4368,
+      "step": 11532
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7220857390748701,
+      "learning_rate": 1.7352038611282485e-06,
+      "loss": 0.5184,
+      "step": 11533
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7569950375840301,
+      "learning_rate": 1.7344226143814341e-06,
+      "loss": 0.506,
+      "step": 11534
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5477283439534922,
+      "learning_rate": 1.733641506639816e-06,
+      "loss": 0.4,
+      "step": 11535
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.640102976398902,
+      "learning_rate": 1.7328605379366391e-06,
+      "loss": 0.4001,
+      "step": 11536
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5938063790849741,
+      "learning_rate": 1.7320797083051506e-06,
+      "loss": 0.4166,
+      "step": 11537
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6173801546358957,
+      "learning_rate": 1.7312990177785848e-06,
+      "loss": 0.4189,
+      "step": 11538
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5318305032781782,
+      "learning_rate": 1.7305184663901741e-06,
+      "loss": 0.4228,
+      "step": 11539
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7169237283214026,
+      "learning_rate": 1.7297380541731461e-06,
+      "loss": 0.4193,
+      "step": 11540
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5883173176085361,
+      "learning_rate": 1.7289577811607179e-06,
+      "loss": 0.4777,
+      "step": 11541
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6366062492789052,
+      "learning_rate": 1.728177647386105e-06,
+      "loss": 0.4791,
+      "step": 11542
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6983595428373511,
+      "learning_rate": 1.7273976528825132e-06,
+      "loss": 0.5244,
+      "step": 11543
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7089669432113576,
+      "learning_rate": 1.726617797683147e-06,
+      "loss": 0.5168,
+      "step": 11544
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7120056115174487,
+      "learning_rate": 1.7258380818211995e-06,
+      "loss": 0.5295,
+      "step": 11545
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7434373498383949,
+      "learning_rate": 1.725058505329864e-06,
+      "loss": 0.483,
+      "step": 11546
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6767554191941588,
+      "learning_rate": 1.7242790682423205e-06,
+      "loss": 0.5294,
+      "step": 11547
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7655589063541344,
+      "learning_rate": 1.723499770591751e-06,
+      "loss": 0.5271,
+      "step": 11548
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6070402806373606,
+      "learning_rate": 1.722720612411325e-06,
+      "loss": 0.5066,
+      "step": 11549
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.706250517347182,
+      "learning_rate": 1.7219415937342115e-06,
+      "loss": 0.5281,
+      "step": 11550
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6880917583980143,
+      "learning_rate": 1.7211627145935655e-06,
+      "loss": 0.4951,
+      "step": 11551
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6322374933541931,
+      "learning_rate": 1.720383975022548e-06,
+      "loss": 0.4784,
+      "step": 11552
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.673863985497878,
+      "learning_rate": 1.7196053750543034e-06,
+      "loss": 0.4706,
+      "step": 11553
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7028437483328559,
+      "learning_rate": 1.7188269147219772e-06,
+      "loss": 0.5012,
+      "step": 11554
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6344949671979806,
+      "learning_rate": 1.7180485940587021e-06,
+      "loss": 0.4543,
+      "step": 11555
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6669204747579709,
+      "learning_rate": 1.7172704130976125e-06,
+      "loss": 0.4921,
+      "step": 11556
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7643036261885616,
+      "learning_rate": 1.7164923718718296e-06,
+      "loss": 0.5387,
+      "step": 11557
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6034249820064262,
+      "learning_rate": 1.7157144704144758e-06,
+      "loss": 0.4326,
+      "step": 11558
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.835448185065465,
+      "learning_rate": 1.71493670875866e-06,
+      "loss": 0.4862,
+      "step": 11559
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6776718149410594,
+      "learning_rate": 1.7141590869374925e-06,
+      "loss": 0.475,
+      "step": 11560
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6357531812001118,
+      "learning_rate": 1.7133816049840708e-06,
+      "loss": 0.4488,
+      "step": 11561
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6449287159431228,
+      "learning_rate": 1.7126042629314937e-06,
+      "loss": 0.4583,
+      "step": 11562
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.650441838256173,
+      "learning_rate": 1.7118270608128446e-06,
+      "loss": 0.4691,
+      "step": 11563
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5524188236511227,
+      "learning_rate": 1.7110499986612133e-06,
+      "loss": 0.4289,
+      "step": 11564
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.600231352496326,
+      "learning_rate": 1.710273076509672e-06,
+      "loss": 0.4067,
+      "step": 11565
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6465559373369545,
+      "learning_rate": 1.7094962943912951e-06,
+      "loss": 0.4678,
+      "step": 11566
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5671943370014317,
+      "learning_rate": 1.7087196523391441e-06,
+      "loss": 0.4676,
+      "step": 11567
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6431900325287857,
+      "learning_rate": 1.707943150386282e-06,
+      "loss": 0.458,
+      "step": 11568
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8405237219699712,
+      "learning_rate": 1.7071667885657585e-06,
+      "loss": 0.514,
+      "step": 11569
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6587106144242421,
+      "learning_rate": 1.706390566910624e-06,
+      "loss": 0.4035,
+      "step": 11570
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6780986998984287,
+      "learning_rate": 1.7056144854539164e-06,
+      "loss": 0.4827,
+      "step": 11571
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5870662835236904,
+      "learning_rate": 1.7048385442286741e-06,
+      "loss": 0.4463,
+      "step": 11572
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.645092100606818,
+      "learning_rate": 1.7040627432679241e-06,
+      "loss": 0.4913,
+      "step": 11573
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7125845990609546,
+      "learning_rate": 1.7032870826046922e-06,
+      "loss": 0.5073,
+      "step": 11574
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.853295875398317,
+      "learning_rate": 1.7025115622719929e-06,
+      "loss": 0.4943,
+      "step": 11575
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7416814189030903,
+      "learning_rate": 1.701736182302839e-06,
+      "loss": 0.6,
+      "step": 11576
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7131880675678445,
+      "learning_rate": 1.7009609427302359e-06,
+      "loss": 0.5232,
+      "step": 11577
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8572876704747294,
+      "learning_rate": 1.7001858435871855e-06,
+      "loss": 0.5248,
+      "step": 11578
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7649702228976412,
+      "learning_rate": 1.699410884906677e-06,
+      "loss": 0.4519,
+      "step": 11579
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5346343170959683,
+      "learning_rate": 1.6986360667217016e-06,
+      "loss": 0.4721,
+      "step": 11580
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.713307221952802,
+      "learning_rate": 1.6978613890652373e-06,
+      "loss": 0.4939,
+      "step": 11581
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7546086763042094,
+      "learning_rate": 1.6970868519702638e-06,
+      "loss": 0.5017,
+      "step": 11582
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6932047398458963,
+      "learning_rate": 1.6963124554697464e-06,
+      "loss": 0.4684,
+      "step": 11583
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6003799855910703,
+      "learning_rate": 1.6955381995966524e-06,
+      "loss": 0.4724,
+      "step": 11584
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6089361341662528,
+      "learning_rate": 1.694764084383936e-06,
+      "loss": 0.425,
+      "step": 11585
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6919833018782288,
+      "learning_rate": 1.6939901098645517e-06,
+      "loss": 0.4615,
+      "step": 11586
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6286409655726857,
+      "learning_rate": 1.6932162760714433e-06,
+      "loss": 0.5113,
+      "step": 11587
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7297549599010655,
+      "learning_rate": 1.6924425830375501e-06,
+      "loss": 0.5702,
+      "step": 11588
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7785206896209446,
+      "learning_rate": 1.691669030795809e-06,
+      "loss": 0.5451,
+      "step": 11589
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6674820695033553,
+      "learning_rate": 1.6908956193791432e-06,
+      "loss": 0.4699,
+      "step": 11590
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7202785789230283,
+      "learning_rate": 1.6901223488204781e-06,
+      "loss": 0.5083,
+      "step": 11591
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6012149512688153,
+      "learning_rate": 1.6893492191527266e-06,
+      "loss": 0.4301,
+      "step": 11592
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6552170568313688,
+      "learning_rate": 1.6885762304087995e-06,
+      "loss": 0.5229,
+      "step": 11593
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7565381505453234,
+      "learning_rate": 1.6878033826216016e-06,
+      "loss": 0.4909,
+      "step": 11594
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.6545339847546898,
+      "learning_rate": 1.6870306758240285e-06,
+      "loss": 0.4842,
+      "step": 11595
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7039311974330968,
+      "learning_rate": 1.686258110048974e-06,
+      "loss": 0.5023,
+      "step": 11596
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6477100130079324,
+      "learning_rate": 1.6854856853293212e-06,
+      "loss": 0.4484,
+      "step": 11597
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.763500242358217,
+      "learning_rate": 1.6847134016979521e-06,
+      "loss": 0.4686,
+      "step": 11598
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.3907942308871668,
+      "learning_rate": 1.683941259187738e-06,
+      "loss": 0.516,
+      "step": 11599
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6250893047220832,
+      "learning_rate": 1.6831692578315483e-06,
+      "loss": 0.4873,
+      "step": 11600
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6227874024729992,
+      "learning_rate": 1.6823973976622455e-06,
+      "loss": 0.533,
+      "step": 11601
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7012333196607653,
+      "learning_rate": 1.6816256787126822e-06,
+      "loss": 0.5406,
+      "step": 11602
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7024384664349265,
+      "learning_rate": 1.6808541010157115e-06,
+      "loss": 0.4305,
+      "step": 11603
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6349009486137714,
+      "learning_rate": 1.680082664604174e-06,
+      "loss": 0.4381,
+      "step": 11604
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6092294486809704,
+      "learning_rate": 1.6793113695109097e-06,
+      "loss": 0.4124,
+      "step": 11605
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6150864069303005,
+      "learning_rate": 1.6785402157687474e-06,
+      "loss": 0.4274,
+      "step": 11606
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6286914796107257,
+      "learning_rate": 1.6777692034105158e-06,
+      "loss": 0.4893,
+      "step": 11607
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7341163536527543,
+      "learning_rate": 1.6769983324690315e-06,
+      "loss": 0.4637,
+      "step": 11608
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6791581357361317,
+      "learning_rate": 1.6762276029771113e-06,
+      "loss": 0.5236,
+      "step": 11609
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.9183999407586751,
+      "learning_rate": 1.6754570149675586e-06,
+      "loss": 0.5465,
+      "step": 11610
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8161592850414057,
+      "learning_rate": 1.6746865684731789e-06,
+      "loss": 0.5096,
+      "step": 11611
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6323710622298931,
+      "learning_rate": 1.6739162635267643e-06,
+      "loss": 0.4427,
+      "step": 11612
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6072191275980428,
+      "learning_rate": 1.6731461001611055e-06,
+      "loss": 0.4738,
+      "step": 11613
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5889060218495732,
+      "learning_rate": 1.672376078408986e-06,
+      "loss": 0.4159,
+      "step": 11614
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6562959582773006,
+      "learning_rate": 1.6716061983031851e-06,
+      "loss": 0.4423,
+      "step": 11615
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6160393703774925,
+      "learning_rate": 1.6708364598764703e-06,
+      "loss": 0.5242,
+      "step": 11616
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6883351946369045,
+      "learning_rate": 1.6700668631616112e-06,
+      "loss": 0.4514,
+      "step": 11617
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6254235523005447,
+      "learning_rate": 1.6692974081913627e-06,
+      "loss": 0.4591,
+      "step": 11618
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7676230268042148,
+      "learning_rate": 1.6685280949984823e-06,
+      "loss": 0.451,
+      "step": 11619
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5594107990452631,
+      "learning_rate": 1.6677589236157126e-06,
+      "loss": 0.4797,
+      "step": 11620
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6960349098074288,
+      "learning_rate": 1.6669898940757996e-06,
+      "loss": 0.5289,
+      "step": 11621
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5472261529464447,
+      "learning_rate": 1.6662210064114742e-06,
+      "loss": 0.4429,
+      "step": 11622
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5724791718624103,
+      "learning_rate": 1.6654522606554684e-06,
+      "loss": 0.4138,
+      "step": 11623
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7021882781477281,
+      "learning_rate": 1.6646836568405033e-06,
+      "loss": 0.4889,
+      "step": 11624
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6197098990816787,
+      "learning_rate": 1.6639151949992966e-06,
+      "loss": 0.4532,
+      "step": 11625
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.688513090820751,
+      "learning_rate": 1.6631468751645591e-06,
+      "loss": 0.5016,
+      "step": 11626
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6051277156718824,
+      "learning_rate": 1.6623786973689981e-06,
+      "loss": 0.4522,
+      "step": 11627
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5521555833063058,
+      "learning_rate": 1.6616106616453082e-06,
+      "loss": 0.4509,
+      "step": 11628
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7601612657721504,
+      "learning_rate": 1.6608427680261868e-06,
+      "loss": 0.5949,
+      "step": 11629
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.827222430387003,
+      "learning_rate": 1.6600750165443163e-06,
+      "loss": 0.5038,
+      "step": 11630
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5697281018012231,
+      "learning_rate": 1.6593074072323812e-06,
+      "loss": 0.4075,
+      "step": 11631
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5691538339339532,
+      "learning_rate": 1.6585399401230528e-06,
+      "loss": 0.4276,
+      "step": 11632
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6779596515561276,
+      "learning_rate": 1.657772615249003e-06,
+      "loss": 0.4723,
+      "step": 11633
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6491495786981641,
+      "learning_rate": 1.6570054326428914e-06,
+      "loss": 0.5221,
+      "step": 11634
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8610529673071755,
+      "learning_rate": 1.656238392337377e-06,
+      "loss": 0.5459,
+      "step": 11635
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7328895451009377,
+      "learning_rate": 1.6554714943651078e-06,
+      "loss": 0.4883,
+      "step": 11636
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6491557838985147,
+      "learning_rate": 1.654704738758731e-06,
+      "loss": 0.4438,
+      "step": 11637
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.660849923400595,
+      "learning_rate": 1.6539381255508802e-06,
+      "loss": 0.4893,
+      "step": 11638
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7055581571554427,
+      "learning_rate": 1.6531716547741945e-06,
+      "loss": 0.5549,
+      "step": 11639
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.4094232215773728,
+      "learning_rate": 1.6524053264612945e-06,
+      "loss": 0.574,
+      "step": 11640
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6809876669518666,
+      "learning_rate": 1.651639140644804e-06,
+      "loss": 0.5384,
+      "step": 11641
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.9028688813525833,
+      "learning_rate": 1.6508730973573338e-06,
+      "loss": 0.4999,
+      "step": 11642
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7819055999206425,
+      "learning_rate": 1.6501071966314952e-06,
+      "loss": 0.4797,
+      "step": 11643
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7903758568525185,
+      "learning_rate": 1.6493414384998873e-06,
+      "loss": 0.4718,
+      "step": 11644
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.1374231005083308,
+      "learning_rate": 1.6485758229951082e-06,
+      "loss": 0.5196,
+      "step": 11645
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8436919990140662,
+      "learning_rate": 1.6478103501497455e-06,
+      "loss": 0.5309,
+      "step": 11646
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.762438017786036,
+      "learning_rate": 1.6470450199963856e-06,
+      "loss": 0.4766,
+      "step": 11647
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.780788225908962,
+      "learning_rate": 1.6462798325676033e-06,
+      "loss": 0.5233,
+      "step": 11648
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6717274026315277,
+      "learning_rate": 1.6455147878959727e-06,
+      "loss": 0.5177,
+      "step": 11649
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8202416257120945,
+      "learning_rate": 1.644749886014057e-06,
+      "loss": 0.5928,
+      "step": 11650
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.671860262074767,
+      "learning_rate": 1.6439851269544165e-06,
+      "loss": 0.4137,
+      "step": 11651
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.62919415610537,
+      "learning_rate": 1.6432205107496069e-06,
+      "loss": 0.4568,
+      "step": 11652
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6233772364002257,
+      "learning_rate": 1.6424560374321713e-06,
+      "loss": 0.4446,
+      "step": 11653
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6564702793842595,
+      "learning_rate": 1.6416917070346534e-06,
+      "loss": 0.5211,
+      "step": 11654
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.756661432533553,
+      "learning_rate": 1.6409275195895892e-06,
+      "loss": 0.5062,
+      "step": 11655
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6899665087428536,
+      "learning_rate": 1.6401634751295047e-06,
+      "loss": 0.5137,
+      "step": 11656
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7044619148972615,
+      "learning_rate": 1.6393995736869267e-06,
+      "loss": 0.477,
+      "step": 11657
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6246292546745654,
+      "learning_rate": 1.6386358152943682e-06,
+      "loss": 0.4564,
+      "step": 11658
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5545051837420987,
+      "learning_rate": 1.6378721999843428e-06,
+      "loss": 0.4251,
+      "step": 11659
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5675673493529451,
+      "learning_rate": 1.6371087277893527e-06,
+      "loss": 0.4294,
+      "step": 11660
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6752120206646379,
+      "learning_rate": 1.6363453987418997e-06,
+      "loss": 0.487,
+      "step": 11661
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7313314388976673,
+      "learning_rate": 1.6355822128744725e-06,
+      "loss": 0.546,
+      "step": 11662
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7567602038825845,
+      "learning_rate": 1.6348191702195593e-06,
+      "loss": 0.5424,
+      "step": 11663
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7669195468100269,
+      "learning_rate": 1.6340562708096419e-06,
+      "loss": 0.5364,
+      "step": 11664
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7715835948915516,
+      "learning_rate": 1.6332935146771917e-06,
+      "loss": 0.5629,
+      "step": 11665
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6390565392200357,
+      "learning_rate": 1.6325309018546798e-06,
+      "loss": 0.458,
+      "step": 11666
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7582914358538736,
+      "learning_rate": 1.6317684323745648e-06,
+      "loss": 0.4995,
+      "step": 11667
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5913005676358329,
+      "learning_rate": 1.6310061062693061e-06,
+      "loss": 0.4314,
+      "step": 11668
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6120843872338068,
+      "learning_rate": 1.6302439235713496e-06,
+      "loss": 0.4709,
+      "step": 11669
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7248412225181978,
+      "learning_rate": 1.6294818843131432e-06,
+      "loss": 0.4963,
+      "step": 11670
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6743556696793921,
+      "learning_rate": 1.6287199885271205e-06,
+      "loss": 0.4737,
+      "step": 11671
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.706868681575387,
+      "learning_rate": 1.6279582362457165e-06,
+      "loss": 0.5278,
+      "step": 11672
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.618794612795796,
+      "learning_rate": 1.6271966275013529e-06,
+      "loss": 0.4242,
+      "step": 11673
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6514651043071733,
+      "learning_rate": 1.626435162326453e-06,
+      "loss": 0.4208,
+      "step": 11674
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6757377179089523,
+      "learning_rate": 1.6256738407534245e-06,
+      "loss": 0.5056,
+      "step": 11675
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.6786881160375586,
+      "learning_rate": 1.624912662814681e-06,
+      "loss": 0.5572,
+      "step": 11676
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7211068813967096,
+      "learning_rate": 1.6241516285426183e-06,
+      "loss": 0.5078,
+      "step": 11677
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6589763639653017,
+      "learning_rate": 1.623390737969635e-06,
+      "loss": 0.4542,
+      "step": 11678
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5719843392074726,
+      "learning_rate": 1.6226299911281163e-06,
+      "loss": 0.4577,
+      "step": 11679
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7336040203804371,
+      "learning_rate": 1.6218693880504482e-06,
+      "loss": 0.5266,
+      "step": 11680
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6448540986105957,
+      "learning_rate": 1.6211089287690034e-06,
+      "loss": 0.4804,
+      "step": 11681
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7955898616096426,
+      "learning_rate": 1.6203486133161555e-06,
+      "loss": 0.5364,
+      "step": 11682
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7468945140372171,
+      "learning_rate": 1.6195884417242663e-06,
+      "loss": 0.4587,
+      "step": 11683
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.989954219237031,
+      "learning_rate": 1.618828414025696e-06,
+      "loss": 0.5095,
+      "step": 11684
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7369868046092862,
+      "learning_rate": 1.6180685302527938e-06,
+      "loss": 0.5236,
+      "step": 11685
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7431057810762721,
+      "learning_rate": 1.6173087904379091e-06,
+      "loss": 0.5428,
+      "step": 11686
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.678718643119146,
+      "learning_rate": 1.6165491946133766e-06,
+      "loss": 0.4309,
+      "step": 11687
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.656017728927584,
+      "learning_rate": 1.6157897428115354e-06,
+      "loss": 0.4556,
+      "step": 11688
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6436879738085989,
+      "learning_rate": 1.6150304350647095e-06,
+      "loss": 0.4862,
+      "step": 11689
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7194150380350957,
+      "learning_rate": 1.6142712714052223e-06,
+      "loss": 0.4799,
+      "step": 11690
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.0333355860260767,
+      "learning_rate": 1.6135122518653868e-06,
+      "loss": 0.4995,
+      "step": 11691
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8593372914682009,
+      "learning_rate": 1.6127533764775143e-06,
+      "loss": 0.5552,
+      "step": 11692
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7344956407632456,
+      "learning_rate": 1.611994645273905e-06,
+      "loss": 0.5142,
+      "step": 11693
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.618078695798422,
+      "learning_rate": 1.6112360582868586e-06,
+      "loss": 0.4296,
+      "step": 11694
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.616708423446725,
+      "learning_rate": 1.6104776155486623e-06,
+      "loss": 0.4659,
+      "step": 11695
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6221525712560697,
+      "learning_rate": 1.6097193170916048e-06,
+      "loss": 0.5269,
+      "step": 11696
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8076474980628993,
+      "learning_rate": 1.6089611629479602e-06,
+      "loss": 0.5579,
+      "step": 11697
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7357107454652001,
+      "learning_rate": 1.6082031531500047e-06,
+      "loss": 0.4817,
+      "step": 11698
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5882972688449865,
+      "learning_rate": 1.60744528773e-06,
+      "loss": 0.4384,
+      "step": 11699
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5990767493900242,
+      "learning_rate": 1.6066875667202093e-06,
+      "loss": 0.4412,
+      "step": 11700
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7346472202613129,
+      "learning_rate": 1.6059299901528847e-06,
+      "loss": 0.4271,
+      "step": 11701
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7010105433833499,
+      "learning_rate": 1.6051725580602762e-06,
+      "loss": 0.4774,
+      "step": 11702
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6691420285863616,
+      "learning_rate": 1.6044152704746224e-06,
+      "loss": 0.4907,
+      "step": 11703
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7181996896352522,
+      "learning_rate": 1.603658127428161e-06,
+      "loss": 0.523,
+      "step": 11704
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.642430804792435,
+      "learning_rate": 1.6029011289531183e-06,
+      "loss": 0.4818,
+      "step": 11705
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7907351539196825,
+      "learning_rate": 1.6021442750817208e-06,
+      "loss": 0.5945,
+      "step": 11706
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7397152098901558,
+      "learning_rate": 1.601387565846182e-06,
+      "loss": 0.5181,
+      "step": 11707
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6363073864898816,
+      "learning_rate": 1.6006310012787156e-06,
+      "loss": 0.4721,
+      "step": 11708
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.9345310392348628,
+      "learning_rate": 1.599874581411523e-06,
+      "loss": 0.5058,
+      "step": 11709
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7084655441965263,
+      "learning_rate": 1.5991183062768057e-06,
+      "loss": 0.4809,
+      "step": 11710
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7529786534217483,
+      "learning_rate": 1.598362175906753e-06,
+      "loss": 0.4894,
+      "step": 11711
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7560265271118195,
+      "learning_rate": 1.5976061903335526e-06,
+      "loss": 0.5485,
+      "step": 11712
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7445073342863389,
+      "learning_rate": 1.5968503495893855e-06,
+      "loss": 0.5085,
+      "step": 11713
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7716263320602863,
+      "learning_rate": 1.5960946537064226e-06,
+      "loss": 0.511,
+      "step": 11714
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5971113743480645,
+      "learning_rate": 1.5953391027168342e-06,
+      "loss": 0.4205,
+      "step": 11715
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6234449235150987,
+      "learning_rate": 1.5945836966527794e-06,
+      "loss": 0.5052,
+      "step": 11716
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7259830708394077,
+      "learning_rate": 1.593828435546414e-06,
+      "loss": 0.4817,
+      "step": 11717
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6201405993008865,
+      "learning_rate": 1.5930733194298897e-06,
+      "loss": 0.4132,
+      "step": 11718
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6548015460817552,
+      "learning_rate": 1.5923183483353455e-06,
+      "loss": 0.4445,
+      "step": 11719
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5760114762894553,
+      "learning_rate": 1.5915635222949212e-06,
+      "loss": 0.4694,
+      "step": 11720
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7039268430095816,
+      "learning_rate": 1.590808841340744e-06,
+      "loss": 0.5053,
+      "step": 11721
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5600014365888779,
+      "learning_rate": 1.5900543055049421e-06,
+      "loss": 0.4094,
+      "step": 11722
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6084876811002675,
+      "learning_rate": 1.5892999148196302e-06,
+      "loss": 0.4377,
+      "step": 11723
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6953852629916997,
+      "learning_rate": 1.5885456693169221e-06,
+      "loss": 0.4994,
+      "step": 11724
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6795727701064831,
+      "learning_rate": 1.5877915690289252e-06,
+      "loss": 0.477,
+      "step": 11725
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5947685028672042,
+      "learning_rate": 1.5870376139877353e-06,
+      "loss": 0.4172,
+      "step": 11726
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7346432079184089,
+      "learning_rate": 1.5862838042254498e-06,
+      "loss": 0.4619,
+      "step": 11727
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8121603884955976,
+      "learning_rate": 1.5855301397741518e-06,
+      "loss": 0.5134,
+      "step": 11728
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.658093977326385,
+      "learning_rate": 1.584776620665927e-06,
+      "loss": 0.4814,
+      "step": 11729
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6034520885958667,
+      "learning_rate": 1.584023246932846e-06,
+      "loss": 0.4365,
+      "step": 11730
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5632148660641273,
+      "learning_rate": 1.5832700186069815e-06,
+      "loss": 0.3963,
+      "step": 11731
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6144137899006358,
+      "learning_rate": 1.5825169357203918e-06,
+      "loss": 0.4759,
+      "step": 11732
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.688181547907297,
+      "learning_rate": 1.5817639983051376e-06,
+      "loss": 0.5318,
+      "step": 11733
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7524833399109788,
+      "learning_rate": 1.5810112063932649e-06,
+      "loss": 0.5195,
+      "step": 11734
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6532304991786136,
+      "learning_rate": 1.5802585600168214e-06,
+      "loss": 0.455,
+      "step": 11735
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5896838061345844,
+      "learning_rate": 1.5795060592078415e-06,
+      "loss": 0.47,
+      "step": 11736
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6407654360264194,
+      "learning_rate": 1.578753703998358e-06,
+      "loss": 0.4353,
+      "step": 11737
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6875393592874859,
+      "learning_rate": 1.578001494420397e-06,
+      "loss": 0.4363,
+      "step": 11738
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6161792859934612,
+      "learning_rate": 1.5772494305059788e-06,
+      "loss": 0.4831,
+      "step": 11739
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6526223812951869,
+      "learning_rate": 1.576497512287113e-06,
+      "loss": 0.5036,
+      "step": 11740
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6491918024259562,
+      "learning_rate": 1.5757457397958097e-06,
+      "loss": 0.4523,
+      "step": 11741
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7540643503895571,
+      "learning_rate": 1.574994113064066e-06,
+      "loss": 0.4948,
+      "step": 11742
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5779330418048424,
+      "learning_rate": 1.5742426321238806e-06,
+      "loss": 0.441,
+      "step": 11743
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7225246613868435,
+      "learning_rate": 1.573491297007237e-06,
+      "loss": 0.453,
+      "step": 11744
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7379107104095686,
+      "learning_rate": 1.5727401077461214e-06,
+      "loss": 0.5373,
+      "step": 11745
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7396805372132097,
+      "learning_rate": 1.5719890643725055e-06,
+      "loss": 0.5861,
+      "step": 11746
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7411962699226632,
+      "learning_rate": 1.5712381669183625e-06,
+      "loss": 0.527,
+      "step": 11747
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7519719020075472,
+      "learning_rate": 1.5704874154156518e-06,
+      "loss": 0.5686,
+      "step": 11748
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.968156842929008,
+      "learning_rate": 1.5697368098963334e-06,
+      "loss": 0.5331,
+      "step": 11749
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7384914180403779,
+      "learning_rate": 1.5689863503923568e-06,
+      "loss": 0.4788,
+      "step": 11750
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5429395082431272,
+      "learning_rate": 1.5682360369356698e-06,
+      "loss": 0.402,
+      "step": 11751
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5750268987020277,
+      "learning_rate": 1.5674858695582063e-06,
+      "loss": 0.4096,
+      "step": 11752
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.654286318930066,
+      "learning_rate": 1.5667358482919022e-06,
+      "loss": 0.438,
+      "step": 11753
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6134140235453149,
+      "learning_rate": 1.56598597316868e-06,
+      "loss": 0.4763,
+      "step": 11754
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7039122258646313,
+      "learning_rate": 1.5652362442204632e-06,
+      "loss": 0.4914,
+      "step": 11755
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6116849157370886,
+      "learning_rate": 1.564486661479162e-06,
+      "loss": 0.4096,
+      "step": 11756
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7165708027256114,
+      "learning_rate": 1.5637372249766869e-06,
+      "loss": 0.4889,
+      "step": 11757
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6299564560444582,
+      "learning_rate": 1.5629879347449362e-06,
+      "loss": 0.4847,
+      "step": 11758
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5888890148812191,
+      "learning_rate": 1.562238790815807e-06,
+      "loss": 0.4118,
+      "step": 11759
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7052670554341612,
+      "learning_rate": 1.5614897932211854e-06,
+      "loss": 0.5085,
+      "step": 11760
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8093611104067335,
+      "learning_rate": 1.5607409419929576e-06,
+      "loss": 0.5263,
+      "step": 11761
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6389577030413439,
+      "learning_rate": 1.559992237162994e-06,
+      "loss": 0.4857,
+      "step": 11762
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5891752639213359,
+      "learning_rate": 1.559243678763171e-06,
+      "loss": 0.456,
+      "step": 11763
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6419411020546483,
+      "learning_rate": 1.5584952668253484e-06,
+      "loss": 0.4509,
+      "step": 11764
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6148366827308623,
+      "learning_rate": 1.5577470013813862e-06,
+      "loss": 0.4905,
+      "step": 11765
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7111504954333165,
+      "learning_rate": 1.5569988824631327e-06,
+      "loss": 0.5505,
+      "step": 11766
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7666176035982654,
+      "learning_rate": 1.556250910102436e-06,
+      "loss": 0.5043,
+      "step": 11767
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5837982535188084,
+      "learning_rate": 1.555503084331132e-06,
+      "loss": 0.4352,
+      "step": 11768
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6010695453524361,
+      "learning_rate": 1.5547554051810565e-06,
+      "loss": 0.4371,
+      "step": 11769
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6264840395814519,
+      "learning_rate": 1.5540078726840319e-06,
+      "loss": 0.4703,
+      "step": 11770
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.68022669008752,
+      "learning_rate": 1.553260486871882e-06,
+      "loss": 0.514,
+      "step": 11771
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7956379103897336,
+      "learning_rate": 1.5525132477764176e-06,
+      "loss": 0.5121,
+      "step": 11772
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7154696553375018,
+      "learning_rate": 1.5517661554294494e-06,
+      "loss": 0.4875,
+      "step": 11773
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6287273317750448,
+      "learning_rate": 1.5510192098627752e-06,
+      "loss": 0.4581,
+      "step": 11774
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6814051416071311,
+      "learning_rate": 1.550272411108192e-06,
+      "loss": 0.478,
+      "step": 11775
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6542467001588551,
+      "learning_rate": 1.5495257591974894e-06,
+      "loss": 0.4654,
+      "step": 11776
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.653667677085657,
+      "learning_rate": 1.548779254162448e-06,
+      "loss": 0.5436,
+      "step": 11777
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.802816849316158,
+      "learning_rate": 1.5480328960348456e-06,
+      "loss": 0.4777,
+      "step": 11778
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5988588495379926,
+      "learning_rate": 1.547286684846453e-06,
+      "loss": 0.4143,
+      "step": 11779
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6680500243335126,
+      "learning_rate": 1.5465406206290318e-06,
+      "loss": 0.5123,
+      "step": 11780
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8222643093206636,
+      "learning_rate": 1.5457947034143417e-06,
+      "loss": 0.5485,
+      "step": 11781
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6696696806370672,
+      "learning_rate": 1.5450489332341317e-06,
+      "loss": 0.4826,
+      "step": 11782
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.771063794677532,
+      "learning_rate": 1.5443033101201498e-06,
+      "loss": 0.5505,
+      "step": 11783
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6418626006642414,
+      "learning_rate": 1.5435578341041313e-06,
+      "loss": 0.4411,
+      "step": 11784
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5483683083025235,
+      "learning_rate": 1.5428125052178128e-06,
+      "loss": 0.4549,
+      "step": 11785
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7608689896695447,
+      "learning_rate": 1.5420673234929162e-06,
+      "loss": 0.523,
+      "step": 11786
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7668921287249512,
+      "learning_rate": 1.541322288961164e-06,
+      "loss": 0.5205,
+      "step": 11787
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8093078290301612,
+      "learning_rate": 1.5405774016542707e-06,
+      "loss": 0.5532,
+      "step": 11788
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8690668151779515,
+      "learning_rate": 1.5398326616039415e-06,
+      "loss": 0.5387,
+      "step": 11789
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6192319007616564,
+      "learning_rate": 1.5390880688418808e-06,
+      "loss": 0.47,
+      "step": 11790
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6493168836358363,
+      "learning_rate": 1.538343623399779e-06,
+      "loss": 0.4736,
+      "step": 11791
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7391252325169428,
+      "learning_rate": 1.5375993253093298e-06,
+      "loss": 0.4758,
+      "step": 11792
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6755645686739995,
+      "learning_rate": 1.5368551746022109e-06,
+      "loss": 0.486,
+      "step": 11793
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7293311495375693,
+      "learning_rate": 1.5361111713101017e-06,
+      "loss": 0.5299,
+      "step": 11794
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7938046028945703,
+      "learning_rate": 1.5353673154646703e-06,
+      "loss": 0.4883,
+      "step": 11795
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7300984710314583,
+      "learning_rate": 1.5346236070975823e-06,
+      "loss": 0.5228,
+      "step": 11796
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8588919596365745,
+      "learning_rate": 1.5338800462404918e-06,
+      "loss": 0.5292,
+      "step": 11797
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7817866249813722,
+      "learning_rate": 1.533136632925053e-06,
+      "loss": 0.5224,
+      "step": 11798
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6312476615578303,
+      "learning_rate": 1.532393367182906e-06,
+      "loss": 0.4807,
+      "step": 11799
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6032819000982859,
+      "learning_rate": 1.5316502490456963e-06,
+      "loss": 0.4443,
+      "step": 11800
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7333570353653186,
+      "learning_rate": 1.5309072785450502e-06,
+      "loss": 0.5031,
+      "step": 11801
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.65520263007275,
+      "learning_rate": 1.5301644557125973e-06,
+      "loss": 0.4434,
+      "step": 11802
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6385059594453464,
+      "learning_rate": 1.5294217805799539e-06,
+      "loss": 0.4637,
+      "step": 11803
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7016805321237948,
+      "learning_rate": 1.5286792531787365e-06,
+      "loss": 0.4837,
+      "step": 11804
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8813445668346723,
+      "learning_rate": 1.5279368735405487e-06,
+      "loss": 0.5593,
+      "step": 11805
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6998779931818744,
+      "learning_rate": 1.5271946416969945e-06,
+      "loss": 0.4976,
+      "step": 11806
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.75638765995227,
+      "learning_rate": 1.5264525576796663e-06,
+      "loss": 0.5477,
+      "step": 11807
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7014983236205438,
+      "learning_rate": 1.5257106215201533e-06,
+      "loss": 0.4773,
+      "step": 11808
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5784815778331148,
+      "learning_rate": 1.5249688332500362e-06,
+      "loss": 0.4077,
+      "step": 11809
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6860376601882198,
+      "learning_rate": 1.5242271929008923e-06,
+      "loss": 0.5176,
+      "step": 11810
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.708522576587283,
+      "learning_rate": 1.523485700504287e-06,
+      "loss": 0.4995,
+      "step": 11811
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8573932136254214,
+      "learning_rate": 1.5227443560917894e-06,
+      "loss": 0.5143,
+      "step": 11812
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8141553424477549,
+      "learning_rate": 1.5220031596949508e-06,
+      "loss": 0.47,
+      "step": 11813
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.799925582347533,
+      "learning_rate": 1.5212621113453252e-06,
+      "loss": 0.5356,
+      "step": 11814
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.735865505739735,
+      "learning_rate": 1.5205212110744538e-06,
+      "loss": 0.5113,
+      "step": 11815
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5881047136717945,
+      "learning_rate": 1.519780458913877e-06,
+      "loss": 0.4616,
+      "step": 11816
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6602627792109644,
+      "learning_rate": 1.519039854895123e-06,
+      "loss": 0.5027,
+      "step": 11817
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8064955582453905,
+      "learning_rate": 1.5182993990497208e-06,
+      "loss": 0.4981,
+      "step": 11818
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6263308430897092,
+      "learning_rate": 1.5175590914091853e-06,
+      "loss": 0.5091,
+      "step": 11819
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7712133292642409,
+      "learning_rate": 1.516818932005033e-06,
+      "loss": 0.5092,
+      "step": 11820
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.589759177810315,
+      "learning_rate": 1.5160789208687665e-06,
+      "loss": 0.4382,
+      "step": 11821
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5582565081809934,
+      "learning_rate": 1.5153390580318888e-06,
+      "loss": 0.438,
+      "step": 11822
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6389886936717366,
+      "learning_rate": 1.5145993435258905e-06,
+      "loss": 0.4795,
+      "step": 11823
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.716650432743136,
+      "learning_rate": 1.5138597773822606e-06,
+      "loss": 0.464,
+      "step": 11824
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6478239783158753,
+      "learning_rate": 1.5131203596324795e-06,
+      "loss": 0.5096,
+      "step": 11825
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6454641410197964,
+      "learning_rate": 1.5123810903080239e-06,
+      "loss": 0.4651,
+      "step": 11826
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6202316761037979,
+      "learning_rate": 1.5116419694403588e-06,
+      "loss": 0.4546,
+      "step": 11827
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6202174516241951,
+      "learning_rate": 1.5109029970609495e-06,
+      "loss": 0.4357,
+      "step": 11828
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6908969314066419,
+      "learning_rate": 1.5101641732012484e-06,
+      "loss": 0.5059,
+      "step": 11829
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7185467599864933,
+      "learning_rate": 1.509425497892708e-06,
+      "loss": 0.513,
+      "step": 11830
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7064557693082774,
+      "learning_rate": 1.5086869711667684e-06,
+      "loss": 0.4665,
+      "step": 11831
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.578798687162687,
+      "learning_rate": 1.50794859305487e-06,
+      "loss": 0.4932,
+      "step": 11832
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6905663030653478,
+      "learning_rate": 1.5072103635884384e-06,
+      "loss": 0.5603,
+      "step": 11833
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7510664241129372,
+      "learning_rate": 1.5064722827989026e-06,
+      "loss": 0.4585,
+      "step": 11834
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6273943909358254,
+      "learning_rate": 1.505734350717676e-06,
+      "loss": 0.4298,
+      "step": 11835
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6419214079985587,
+      "learning_rate": 1.504996567376172e-06,
+      "loss": 0.4343,
+      "step": 11836
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.678229937919406,
+      "learning_rate": 1.5042589328057977e-06,
+      "loss": 0.4462,
+      "step": 11837
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6283421918971349,
+      "learning_rate": 1.5035214470379478e-06,
+      "loss": 0.455,
+      "step": 11838
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6304615540000009,
+      "learning_rate": 1.5027841101040186e-06,
+      "loss": 0.4386,
+      "step": 11839
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.703904198555036,
+      "learning_rate": 1.5020469220353928e-06,
+      "loss": 0.4178,
+      "step": 11840
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6089485199679933,
+      "learning_rate": 1.5013098828634514e-06,
+      "loss": 0.4245,
+      "step": 11841
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7898261139824619,
+      "learning_rate": 1.50057299261957e-06,
+      "loss": 0.5052,
+      "step": 11842
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7025556079504152,
+      "learning_rate": 1.4998362513351118e-06,
+      "loss": 0.5037,
+      "step": 11843
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5482683859555755,
+      "learning_rate": 1.4990996590414408e-06,
+      "loss": 0.438,
+      "step": 11844
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5988178399422383,
+      "learning_rate": 1.4983632157699091e-06,
+      "loss": 0.4193,
+      "step": 11845
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6101592647269215,
+      "learning_rate": 1.4976269215518667e-06,
+      "loss": 0.4284,
+      "step": 11846
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6584350770773381,
+      "learning_rate": 1.4968907764186535e-06,
+      "loss": 0.4462,
+      "step": 11847
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.644409796959749,
+      "learning_rate": 1.4961547804016046e-06,
+      "loss": 0.4285,
+      "step": 11848
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7596332590127007,
+      "learning_rate": 1.4954189335320524e-06,
+      "loss": 0.5522,
+      "step": 11849
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7126415012322277,
+      "learning_rate": 1.4946832358413154e-06,
+      "loss": 0.4777,
+      "step": 11850
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8145319375764453,
+      "learning_rate": 1.4939476873607129e-06,
+      "loss": 0.5306,
+      "step": 11851
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5730804762496607,
+      "learning_rate": 1.4932122881215522e-06,
+      "loss": 0.4582,
+      "step": 11852
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7628296180292785,
+      "learning_rate": 1.49247703815514e-06,
+      "loss": 0.5631,
+      "step": 11853
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7769824708820705,
+      "learning_rate": 1.49174193749277e-06,
+      "loss": 0.5013,
+      "step": 11854
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6436061963935515,
+      "learning_rate": 1.4910069861657367e-06,
+      "loss": 0.4509,
+      "step": 11855
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6284176954190089,
+      "learning_rate": 1.4902721842053215e-06,
+      "loss": 0.4675,
+      "step": 11856
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.857460364630511,
+      "learning_rate": 1.489537531642805e-06,
+      "loss": 0.5037,
+      "step": 11857
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.4153407418352368,
+      "learning_rate": 1.4888030285094569e-06,
+      "loss": 0.5133,
+      "step": 11858
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7606255532426416,
+      "learning_rate": 1.4880686748365453e-06,
+      "loss": 0.5251,
+      "step": 11859
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7559046403360551,
+      "learning_rate": 1.4873344706553255e-06,
+      "loss": 0.4628,
+      "step": 11860
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7370367427270319,
+      "learning_rate": 1.4866004159970527e-06,
+      "loss": 0.5052,
+      "step": 11861
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8421025089555793,
+      "learning_rate": 1.4858665108929732e-06,
+      "loss": 0.5515,
+      "step": 11862
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7308373939186672,
+      "learning_rate": 1.4851327553743283e-06,
+      "loss": 0.528,
+      "step": 11863
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7749088901483504,
+      "learning_rate": 1.484399149472348e-06,
+      "loss": 0.4869,
+      "step": 11864
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5833079333836493,
+      "learning_rate": 1.4836656932182635e-06,
+      "loss": 0.4123,
+      "step": 11865
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7145071771547642,
+      "learning_rate": 1.4829323866432927e-06,
+      "loss": 0.5104,
+      "step": 11866
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7416386903058149,
+      "learning_rate": 1.4821992297786524e-06,
+      "loss": 0.5,
+      "step": 11867
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.717791106770553,
+      "learning_rate": 1.4814662226555482e-06,
+      "loss": 0.4784,
+      "step": 11868
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7302412626467278,
+      "learning_rate": 1.4807333653051848e-06,
+      "loss": 0.5481,
+      "step": 11869
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7592125182075823,
+      "learning_rate": 1.4800006577587545e-06,
+      "loss": 0.5285,
+      "step": 11870
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6763326314552282,
+      "learning_rate": 1.4792681000474496e-06,
+      "loss": 0.4864,
+      "step": 11871
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6598942039187701,
+      "learning_rate": 1.4785356922024491e-06,
+      "loss": 0.4778,
+      "step": 11872
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.575117547414073,
+      "learning_rate": 1.4778034342549313e-06,
+      "loss": 0.4069,
+      "step": 11873
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.643936184193435,
+      "learning_rate": 1.4770713262360664e-06,
+      "loss": 0.47,
+      "step": 11874
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5687847044527354,
+      "learning_rate": 1.476339368177019e-06,
+      "loss": 0.4304,
+      "step": 11875
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6134730889666818,
+      "learning_rate": 1.4756075601089426e-06,
+      "loss": 0.4582,
+      "step": 11876
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.574258200621919,
+      "learning_rate": 1.4748759020629916e-06,
+      "loss": 0.4603,
+      "step": 11877
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7289514713332602,
+      "learning_rate": 1.4741443940703076e-06,
+      "loss": 0.4448,
+      "step": 11878
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6754957264099952,
+      "learning_rate": 1.4734130361620308e-06,
+      "loss": 0.4588,
+      "step": 11879
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5779789544473176,
+      "learning_rate": 1.4726818283692907e-06,
+      "loss": 0.4729,
+      "step": 11880
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7579048188425673,
+      "learning_rate": 1.4719507707232145e-06,
+      "loss": 0.5336,
+      "step": 11881
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7298442253391793,
+      "learning_rate": 1.471219863254919e-06,
+      "loss": 0.5254,
+      "step": 11882
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6527989485826166,
+      "learning_rate": 1.4704891059955184e-06,
+      "loss": 0.4636,
+      "step": 11883
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6276337234324153,
+      "learning_rate": 1.469758498976117e-06,
+      "loss": 0.4765,
+      "step": 11884
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7632291157898848,
+      "learning_rate": 1.4690280422278164e-06,
+      "loss": 0.5636,
+      "step": 11885
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7063143936721762,
+      "learning_rate": 1.4682977357817057e-06,
+      "loss": 0.5329,
+      "step": 11886
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6768998483318774,
+      "learning_rate": 1.4675675796688777e-06,
+      "loss": 0.4854,
+      "step": 11887
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6882104109259862,
+      "learning_rate": 1.4668375739204083e-06,
+      "loss": 0.4776,
+      "step": 11888
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7901680166928836,
+      "learning_rate": 1.4661077185673745e-06,
+      "loss": 0.578,
+      "step": 11889
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8936742457878368,
+      "learning_rate": 1.4653780136408407e-06,
+      "loss": 0.5461,
+      "step": 11890
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7593327976144598,
+      "learning_rate": 1.4646484591718717e-06,
+      "loss": 0.4654,
+      "step": 11891
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6599733636141957,
+      "learning_rate": 1.4639190551915189e-06,
+      "loss": 0.4189,
+      "step": 11892
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7696281045056309,
+      "learning_rate": 1.4631898017308338e-06,
+      "loss": 0.4771,
+      "step": 11893
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8125166737913756,
+      "learning_rate": 1.462460698820855e-06,
+      "loss": 0.5116,
+      "step": 11894
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.760748975049582,
+      "learning_rate": 1.4617317464926217e-06,
+      "loss": 0.5034,
+      "step": 11895
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5917640577108612,
+      "learning_rate": 1.4610029447771595e-06,
+      "loss": 0.4541,
+      "step": 11896
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6254715694753664,
+      "learning_rate": 1.4602742937054942e-06,
+      "loss": 0.4268,
+      "step": 11897
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6612754649362964,
+      "learning_rate": 1.45954579330864e-06,
+      "loss": 0.5101,
+      "step": 11898
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7639911655799347,
+      "learning_rate": 1.4588174436176073e-06,
+      "loss": 0.5033,
+      "step": 11899
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6107291144120331,
+      "learning_rate": 1.4580892446634015e-06,
+      "loss": 0.4513,
+      "step": 11900
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7302142075022269,
+      "learning_rate": 1.4573611964770168e-06,
+      "loss": 0.4629,
+      "step": 11901
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6985277886307565,
+      "learning_rate": 1.4566332990894456e-06,
+      "loss": 0.4996,
+      "step": 11902
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7858628190559802,
+      "learning_rate": 1.455905552531673e-06,
+      "loss": 0.53,
+      "step": 11903
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6550350598902999,
+      "learning_rate": 1.455177956834674e-06,
+      "loss": 0.484,
+      "step": 11904
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7473365611735782,
+      "learning_rate": 1.4544505120294239e-06,
+      "loss": 0.5013,
+      "step": 11905
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7978067914947106,
+      "learning_rate": 1.4537232181468835e-06,
+      "loss": 0.5719,
+      "step": 11906
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6288148890051874,
+      "learning_rate": 1.4529960752180156e-06,
+      "loss": 0.4793,
+      "step": 11907
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6194600075601276,
+      "learning_rate": 1.4522690832737679e-06,
+      "loss": 0.4572,
+      "step": 11908
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7936754033642945,
+      "learning_rate": 1.4515422423450902e-06,
+      "loss": 0.5541,
+      "step": 11909
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6286049158787025,
+      "learning_rate": 1.4508155524629186e-06,
+      "loss": 0.4746,
+      "step": 11910
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7241940144325015,
+      "learning_rate": 1.4500890136581875e-06,
+      "loss": 0.516,
+      "step": 11911
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5373397971462176,
+      "learning_rate": 1.4493626259618242e-06,
+      "loss": 0.3974,
+      "step": 11912
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6185760583103315,
+      "learning_rate": 1.4486363894047468e-06,
+      "loss": 0.4874,
+      "step": 11913
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7521964921008295,
+      "learning_rate": 1.447910304017871e-06,
+      "loss": 0.4835,
+      "step": 11914
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6329308429095662,
+      "learning_rate": 1.4471843698321009e-06,
+      "loss": 0.4769,
+      "step": 11915
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7325882968022589,
+      "learning_rate": 1.446458586878341e-06,
+      "loss": 0.4859,
+      "step": 11916
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6366233076546924,
+      "learning_rate": 1.4457329551874816e-06,
+      "loss": 0.4526,
+      "step": 11917
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6325091054991279,
+      "learning_rate": 1.4450074747904142e-06,
+      "loss": 0.4685,
+      "step": 11918
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.554474044348401,
+      "learning_rate": 1.4442821457180167e-06,
+      "loss": 0.4191,
+      "step": 11919
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5816684748831272,
+      "learning_rate": 1.443556968001168e-06,
+      "loss": 0.4071,
+      "step": 11920
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8849647868166594,
+      "learning_rate": 1.4428319416707327e-06,
+      "loss": 0.4318,
+      "step": 11921
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.748242272080491,
+      "learning_rate": 1.4421070667575754e-06,
+      "loss": 0.5431,
+      "step": 11922
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9153034293356291,
+      "learning_rate": 1.4413823432925483e-06,
+      "loss": 0.5074,
+      "step": 11923
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.622396773485618,
+      "learning_rate": 1.4406577713065061e-06,
+      "loss": 0.4936,
+      "step": 11924
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7407669943325329,
+      "learning_rate": 1.4399333508302871e-06,
+      "loss": 0.5255,
+      "step": 11925
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7967093278333663,
+      "learning_rate": 1.4392090818947308e-06,
+      "loss": 0.5271,
+      "step": 11926
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6152531122051506,
+      "learning_rate": 1.4384849645306631e-06,
+      "loss": 0.4728,
+      "step": 11927
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8028136064244589,
+      "learning_rate": 1.4377609987689117e-06,
+      "loss": 0.551,
+      "step": 11928
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6163416288227633,
+      "learning_rate": 1.437037184640289e-06,
+      "loss": 0.4352,
+      "step": 11929
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9569357815659029,
+      "learning_rate": 1.4363135221756097e-06,
+      "loss": 0.4821,
+      "step": 11930
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.649481211374458,
+      "learning_rate": 1.4355900114056743e-06,
+      "loss": 0.4788,
+      "step": 11931
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.673899160816937,
+      "learning_rate": 1.4348666523612837e-06,
+      "loss": 0.4911,
+      "step": 11932
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8855748910508573,
+      "learning_rate": 1.434143445073225e-06,
+      "loss": 0.5546,
+      "step": 11933
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7517144380153586,
+      "learning_rate": 1.433420389572287e-06,
+      "loss": 0.491,
+      "step": 11934
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6736783602357297,
+      "learning_rate": 1.4326974858892429e-06,
+      "loss": 0.481,
+      "step": 11935
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7518184370772593,
+      "learning_rate": 1.43197473405487e-06,
+      "loss": 0.4954,
+      "step": 11936
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.684843127693918,
+      "learning_rate": 1.4312521340999297e-06,
+      "loss": 0.4079,
+      "step": 11937
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7028455966540011,
+      "learning_rate": 1.4305296860551831e-06,
+      "loss": 0.4909,
+      "step": 11938
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7536326138616248,
+      "learning_rate": 1.4298073899513803e-06,
+      "loss": 0.4917,
+      "step": 11939
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5923788123906366,
+      "learning_rate": 1.4290852458192695e-06,
+      "loss": 0.4344,
+      "step": 11940
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5748647703219977,
+      "learning_rate": 1.428363253689587e-06,
+      "loss": 0.4234,
+      "step": 11941
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5871307327750466,
+      "learning_rate": 1.4276414135930695e-06,
+      "loss": 0.4203,
+      "step": 11942
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6646586427845788,
+      "learning_rate": 1.4269197255604395e-06,
+      "loss": 0.4819,
+      "step": 11943
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7639983366964239,
+      "learning_rate": 1.4261981896224209e-06,
+      "loss": 0.4926,
+      "step": 11944
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5927902522432147,
+      "learning_rate": 1.4254768058097228e-06,
+      "loss": 0.4491,
+      "step": 11945
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7494902058550442,
+      "learning_rate": 1.4247555741530566e-06,
+      "loss": 0.4972,
+      "step": 11946
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.824565801744356,
+      "learning_rate": 1.4240344946831192e-06,
+      "loss": 0.4909,
+      "step": 11947
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6245604221413591,
+      "learning_rate": 1.423313567430606e-06,
+      "loss": 0.4654,
+      "step": 11948
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5799736849700223,
+      "learning_rate": 1.422592792426205e-06,
+      "loss": 0.4471,
+      "step": 11949
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7624688173524956,
+      "learning_rate": 1.4218721697005984e-06,
+      "loss": 0.4586,
+      "step": 11950
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6614371141124121,
+      "learning_rate": 1.421151699284458e-06,
+      "loss": 0.5309,
+      "step": 11951
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.746829092014637,
+      "learning_rate": 1.4204313812084552e-06,
+      "loss": 0.4954,
+      "step": 11952
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6761957546420123,
+      "learning_rate": 1.4197112155032484e-06,
+      "loss": 0.504,
+      "step": 11953
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6896137298686484,
+      "learning_rate": 1.418991202199495e-06,
+      "loss": 0.4449,
+      "step": 11954
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5946682175975541,
+      "learning_rate": 1.4182713413278421e-06,
+      "loss": 0.4231,
+      "step": 11955
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6378199333418438,
+      "learning_rate": 1.4175516329189336e-06,
+      "loss": 0.5385,
+      "step": 11956
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8033376375797316,
+      "learning_rate": 1.4168320770034027e-06,
+      "loss": 0.5389,
+      "step": 11957
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8003877712300571,
+      "learning_rate": 1.4161126736118818e-06,
+      "loss": 0.5532,
+      "step": 11958
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7013408170494367,
+      "learning_rate": 1.4153934227749906e-06,
+      "loss": 0.5007,
+      "step": 11959
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5777901475393556,
+      "learning_rate": 1.4146743245233463e-06,
+      "loss": 0.445,
+      "step": 11960
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.60618280587829,
+      "learning_rate": 1.413955378887561e-06,
+      "loss": 0.4237,
+      "step": 11961
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6977927192262073,
+      "learning_rate": 1.4132365858982339e-06,
+      "loss": 0.4372,
+      "step": 11962
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5529689275852305,
+      "learning_rate": 1.4125179455859634e-06,
+      "loss": 0.4435,
+      "step": 11963
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6465283117409542,
+      "learning_rate": 1.4117994579813421e-06,
+      "loss": 0.4524,
+      "step": 11964
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6324985273586289,
+      "learning_rate": 1.4110811231149495e-06,
+      "loss": 0.4954,
+      "step": 11965
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6556563841419782,
+      "learning_rate": 1.410362941017367e-06,
+      "loss": 0.4788,
+      "step": 11966
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6874829131602086,
+      "learning_rate": 1.4096449117191619e-06,
+      "loss": 0.3874,
+      "step": 11967
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.659648034006389,
+      "learning_rate": 1.408927035250901e-06,
+      "loss": 0.4688,
+      "step": 11968
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.760973846093577,
+      "learning_rate": 1.408209311643139e-06,
+      "loss": 0.4835,
+      "step": 11969
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7310589069948038,
+      "learning_rate": 1.4074917409264304e-06,
+      "loss": 0.4855,
+      "step": 11970
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.683464203817682,
+      "learning_rate": 1.4067743231313174e-06,
+      "loss": 0.5,
+      "step": 11971
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.596613413010919,
+      "learning_rate": 1.4060570582883388e-06,
+      "loss": 0.4422,
+      "step": 11972
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6278333890724107,
+      "learning_rate": 1.4053399464280282e-06,
+      "loss": 0.4571,
+      "step": 11973
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7844696714547017,
+      "learning_rate": 1.4046229875809076e-06,
+      "loss": 0.5401,
+      "step": 11974
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6894544135343352,
+      "learning_rate": 1.403906181777499e-06,
+      "loss": 0.5196,
+      "step": 11975
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.715277767498412,
+      "learning_rate": 1.4031895290483112e-06,
+      "loss": 0.5197,
+      "step": 11976
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8171187872423455,
+      "learning_rate": 1.4024730294238525e-06,
+      "loss": 0.4879,
+      "step": 11977
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6373667151281672,
+      "learning_rate": 1.4017566829346196e-06,
+      "loss": 0.4866,
+      "step": 11978
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5903546033363626,
+      "learning_rate": 1.401040489611108e-06,
+      "loss": 0.4915,
+      "step": 11979
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6854696279920867,
+      "learning_rate": 1.4003244494838003e-06,
+      "loss": 0.5144,
+      "step": 11980
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5707116265821914,
+      "learning_rate": 1.3996085625831801e-06,
+      "loss": 0.4466,
+      "step": 11981
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.559962507230934,
+      "learning_rate": 1.3988928289397158e-06,
+      "loss": 0.4623,
+      "step": 11982
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6003526781341626,
+      "learning_rate": 1.3981772485838785e-06,
+      "loss": 0.4422,
+      "step": 11983
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6364568543294888,
+      "learning_rate": 1.397461821546124e-06,
+      "loss": 0.4571,
+      "step": 11984
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.752513548557085,
+      "learning_rate": 1.3967465478569075e-06,
+      "loss": 0.4927,
+      "step": 11985
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6143365125200207,
+      "learning_rate": 1.3960314275466768e-06,
+      "loss": 0.4491,
+      "step": 11986
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7371996943701631,
+      "learning_rate": 1.3953164606458724e-06,
+      "loss": 0.5085,
+      "step": 11987
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.776388144686789,
+      "learning_rate": 1.3946016471849267e-06,
+      "loss": 0.5261,
+      "step": 11988
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5953005002791292,
+      "learning_rate": 1.3938869871942683e-06,
+      "loss": 0.4367,
+      "step": 11989
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6058115728198914,
+      "learning_rate": 1.3931724807043161e-06,
+      "loss": 0.4764,
+      "step": 11990
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.630541521661115,
+      "learning_rate": 1.3924581277454873e-06,
+      "loss": 0.4654,
+      "step": 11991
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7763509580356418,
+      "learning_rate": 1.3917439283481865e-06,
+      "loss": 0.5119,
+      "step": 11992
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6724843055458318,
+      "learning_rate": 1.3910298825428175e-06,
+      "loss": 0.482,
+      "step": 11993
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6716993553319668,
+      "learning_rate": 1.3903159903597728e-06,
+      "loss": 0.4846,
+      "step": 11994
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9784978355260202,
+      "learning_rate": 1.3896022518294428e-06,
+      "loss": 0.5521,
+      "step": 11995
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7067033772717441,
+      "learning_rate": 1.388888666982206e-06,
+      "loss": 0.5196,
+      "step": 11996
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7225292070001212,
+      "learning_rate": 1.388175235848439e-06,
+      "loss": 0.4127,
+      "step": 11997
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5543762154454209,
+      "learning_rate": 1.387461958458511e-06,
+      "loss": 0.3994,
+      "step": 11998
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.599151813254321,
+      "learning_rate": 1.386748834842785e-06,
+      "loss": 0.4379,
+      "step": 11999
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5865043355783855,
+      "learning_rate": 1.386035865031613e-06,
+      "loss": 0.4301,
+      "step": 12000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 15698,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "total_flos": 5811613688332288.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}