diff --git "a/parm/trainer_state.json" "b/parm/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/parm/trainer_state.json"
@@ -0,0 +1,35280 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999006852716258,
+  "eval_steps": 500,
+  "global_step": 5034,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00019862945674843578,
+      "grad_norm": 33.76269020949264,
+      "learning_rate": 6.578947368421053e-08,
+      "loss": 0.2359,
+      "step": 1
+    },
+    {
+      "epoch": 0.00039725891349687157,
+      "grad_norm": 34.06001813445965,
+      "learning_rate": 1.3157894736842107e-07,
+      "loss": 0.2755,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005958883702453074,
+      "grad_norm": 35.82083527820956,
+      "learning_rate": 1.9736842105263157e-07,
+      "loss": 0.285,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007945178269937431,
+      "grad_norm": 31.578743559628386,
+      "learning_rate": 2.6315789473684213e-07,
+      "loss": 0.2475,
+      "step": 4
+    },
+    {
+      "epoch": 0.000993147283742179,
+      "grad_norm": 30.615127259734468,
+      "learning_rate": 3.2894736842105264e-07,
+      "loss": 0.2229,
+      "step": 5
+    },
+    {
+      "epoch": 0.0011917767404906149,
+      "grad_norm": 44.64811115111308,
+      "learning_rate": 3.9473684210526315e-07,
+      "loss": 0.3314,
+      "step": 6
+    },
+    {
+      "epoch": 0.0013904061972390505,
+      "grad_norm": 44.34169533744506,
+      "learning_rate": 4.605263157894737e-07,
+      "loss": 0.2423,
+      "step": 7
+    },
+    {
+      "epoch": 0.0015890356539874863,
+      "grad_norm": 40.11109275785226,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 0.2239,
+      "step": 8
+    },
+    {
+      "epoch": 0.001787665110735922,
+      "grad_norm": 35.00420920021999,
+      "learning_rate": 5.921052631578947e-07,
+      "loss": 0.2018,
+      "step": 9
+    },
+    {
+      "epoch": 0.001986294567484358,
+      "grad_norm": 29.808999667191497,
+      "learning_rate": 6.578947368421053e-07,
+      "loss": 0.107,
+      "step": 10
+    },
+    {
+      "epoch": 0.0021849240242327937,
+      "grad_norm": 17.62819439139421,
+      "learning_rate": 7.236842105263158e-07,
+      "loss": 0.0941,
+      "step": 11
+    },
+    {
+      "epoch": 0.0023835534809812297,
+      "grad_norm": 12.95448401579392,
+      "learning_rate": 7.894736842105263e-07,
+      "loss": 0.0777,
+      "step": 12
+    },
+    {
+      "epoch": 0.0025821829377296653,
+      "grad_norm": 6.615340288787496,
+      "learning_rate": 8.55263157894737e-07,
+      "loss": 0.0557,
+      "step": 13
+    },
+    {
+      "epoch": 0.002780812394478101,
+      "grad_norm": 8.20493935951563,
+      "learning_rate": 9.210526315789474e-07,
+      "loss": 0.0618,
+      "step": 14
+    },
+    {
+      "epoch": 0.002979441851226537,
+      "grad_norm": 6.4854536788019495,
+      "learning_rate": 9.86842105263158e-07,
+      "loss": 0.0597,
+      "step": 15
+    },
+    {
+      "epoch": 0.0031780713079749725,
+      "grad_norm": 5.948598869884694,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 0.057,
+      "step": 16
+    },
+    {
+      "epoch": 0.0033767007647234086,
+      "grad_norm": 1.505342665823492,
+      "learning_rate": 1.118421052631579e-06,
+      "loss": 0.0389,
+      "step": 17
+    },
+    {
+      "epoch": 0.003575330221471844,
+      "grad_norm": 4.357709266078771,
+      "learning_rate": 1.1842105263157894e-06,
+      "loss": 0.0477,
+      "step": 18
+    },
+    {
+      "epoch": 0.00377395967822028,
+      "grad_norm": 4.173575780842314,
+      "learning_rate": 1.25e-06,
+      "loss": 0.0425,
+      "step": 19
+    },
+    {
+      "epoch": 0.003972589134968716,
+      "grad_norm": 4.034253079966795,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 0.0433,
+      "step": 20
+    },
+    {
+      "epoch": 0.004171218591717152,
+      "grad_norm": 3.815661808462217,
+      "learning_rate": 1.3815789473684212e-06,
+      "loss": 0.0417,
+      "step": 21
+    },
+    {
+      "epoch": 0.004369848048465587,
+      "grad_norm": 1.9849545321722113,
+      "learning_rate": 1.4473684210526317e-06,
+      "loss": 0.0349,
+      "step": 22
+    },
+    {
+      "epoch": 0.004568477505214023,
+      "grad_norm": 0.704960225012627,
+      "learning_rate": 1.5131578947368421e-06,
+      "loss": 0.0355,
+      "step": 23
+    },
+    {
+      "epoch": 0.0047671069619624595,
+      "grad_norm": 2.738263124201555,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 0.0351,
+      "step": 24
+    },
+    {
+      "epoch": 0.004965736418710895,
+      "grad_norm": 3.7531322321385994,
+      "learning_rate": 1.6447368421052635e-06,
+      "loss": 0.0337,
+      "step": 25
+    },
+    {
+      "epoch": 0.005164365875459331,
+      "grad_norm": 3.3650206269293643,
+      "learning_rate": 1.710526315789474e-06,
+      "loss": 0.0403,
+      "step": 26
+    },
+    {
+      "epoch": 0.005362995332207766,
+      "grad_norm": 1.1524716199784135,
+      "learning_rate": 1.7763157894736844e-06,
+      "loss": 0.0188,
+      "step": 27
+    },
+    {
+      "epoch": 0.005561624788956202,
+      "grad_norm": 0.9219916829535164,
+      "learning_rate": 1.8421052631578948e-06,
+      "loss": 0.0292,
+      "step": 28
+    },
+    {
+      "epoch": 0.005760254245704638,
+      "grad_norm": 1.5976055474296411,
+      "learning_rate": 1.9078947368421057e-06,
+      "loss": 0.0311,
+      "step": 29
+    },
+    {
+      "epoch": 0.005958883702453074,
+      "grad_norm": 1.7121779218172795,
+      "learning_rate": 1.973684210526316e-06,
+      "loss": 0.0267,
+      "step": 30
+    },
+    {
+      "epoch": 0.0061575131592015095,
+      "grad_norm": 1.4089365136243617,
+      "learning_rate": 2.0394736842105266e-06,
+      "loss": 0.0271,
+      "step": 31
+    },
+    {
+      "epoch": 0.006356142615949945,
+      "grad_norm": 1.2326619516469113,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 0.0213,
+      "step": 32
+    },
+    {
+      "epoch": 0.0065547720726983815,
+      "grad_norm": 2.5437119531443457,
+      "learning_rate": 2.1710526315789475e-06,
+      "loss": 0.0358,
+      "step": 33
+    },
+    {
+      "epoch": 0.006753401529446817,
+      "grad_norm": 3.097227411879207,
+      "learning_rate": 2.236842105263158e-06,
+      "loss": 0.026,
+      "step": 34
+    },
+    {
+      "epoch": 0.006952030986195253,
+      "grad_norm": 1.5225819387838984,
+      "learning_rate": 2.3026315789473684e-06,
+      "loss": 0.0352,
+      "step": 35
+    },
+    {
+      "epoch": 0.007150660442943688,
+      "grad_norm": 1.9028774569921925,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 0.0249,
+      "step": 36
+    },
+    {
+      "epoch": 0.007349289899692124,
+      "grad_norm": 1.3413799412407437,
+      "learning_rate": 2.4342105263157898e-06,
+      "loss": 0.023,
+      "step": 37
+    },
+    {
+      "epoch": 0.00754791935644056,
+      "grad_norm": 0.7403867166496259,
+      "learning_rate": 2.5e-06,
+      "loss": 0.0179,
+      "step": 38
+    },
+    {
+      "epoch": 0.007746548813188996,
+      "grad_norm": 1.5339334077472897,
+      "learning_rate": 2.565789473684211e-06,
+      "loss": 0.026,
+      "step": 39
+    },
+    {
+      "epoch": 0.007945178269937432,
+      "grad_norm": 1.4657589777219318,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 0.0195,
+      "step": 40
+    },
+    {
+      "epoch": 0.008143807726685867,
+      "grad_norm": 1.3319716700261164,
+      "learning_rate": 2.697368421052632e-06,
+      "loss": 0.0291,
+      "step": 41
+    },
+    {
+      "epoch": 0.008342437183434304,
+      "grad_norm": 1.1874773600556614,
+      "learning_rate": 2.7631578947368424e-06,
+      "loss": 0.0259,
+      "step": 42
+    },
+    {
+      "epoch": 0.008541066640182738,
+      "grad_norm": 0.6674176354529916,
+      "learning_rate": 2.828947368421053e-06,
+      "loss": 0.0203,
+      "step": 43
+    },
+    {
+      "epoch": 0.008739696096931175,
+      "grad_norm": 1.1733771223638585,
+      "learning_rate": 2.8947368421052634e-06,
+      "loss": 0.026,
+      "step": 44
+    },
+    {
+      "epoch": 0.008938325553679611,
+      "grad_norm": 1.7753106520406625,
+      "learning_rate": 2.960526315789474e-06,
+      "loss": 0.0247,
+      "step": 45
+    },
+    {
+      "epoch": 0.009136955010428046,
+      "grad_norm": 0.9432853830345364,
+      "learning_rate": 3.0263157894736843e-06,
+      "loss": 0.0257,
+      "step": 46
+    },
+    {
+      "epoch": 0.009335584467176482,
+      "grad_norm": 3.5238268219753857,
+      "learning_rate": 3.092105263157895e-06,
+      "loss": 0.024,
+      "step": 47
+    },
+    {
+      "epoch": 0.009534213923924919,
+      "grad_norm": 4.0482869582706975,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 0.0279,
+      "step": 48
+    },
+    {
+      "epoch": 0.009732843380673354,
+      "grad_norm": 1.9309599368634551,
+      "learning_rate": 3.223684210526316e-06,
+      "loss": 0.0307,
+      "step": 49
+    },
+    {
+      "epoch": 0.00993147283742179,
+      "grad_norm": 2.2369624906638355,
+      "learning_rate": 3.289473684210527e-06,
+      "loss": 0.0192,
+      "step": 50
+    },
+    {
+      "epoch": 0.010130102294170225,
+      "grad_norm": 1.9485240974125297,
+      "learning_rate": 3.355263157894737e-06,
+      "loss": 0.0286,
+      "step": 51
+    },
+    {
+      "epoch": 0.010328731750918661,
+      "grad_norm": 2.7566490930720473,
+      "learning_rate": 3.421052631578948e-06,
+      "loss": 0.0327,
+      "step": 52
+    },
+    {
+      "epoch": 0.010527361207667098,
+      "grad_norm": 0.9111515455287649,
+      "learning_rate": 3.486842105263158e-06,
+      "loss": 0.0266,
+      "step": 53
+    },
+    {
+      "epoch": 0.010725990664415532,
+      "grad_norm": 1.0182592431220034,
+      "learning_rate": 3.5526315789473687e-06,
+      "loss": 0.0159,
+      "step": 54
+    },
+    {
+      "epoch": 0.010924620121163969,
+      "grad_norm": 1.3689812247740223,
+      "learning_rate": 3.618421052631579e-06,
+      "loss": 0.0236,
+      "step": 55
+    },
+    {
+      "epoch": 0.011123249577912404,
+      "grad_norm": 1.415669213506178,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 0.0192,
+      "step": 56
+    },
+    {
+      "epoch": 0.01132187903466084,
+      "grad_norm": 2.727013850743499,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.0271,
+      "step": 57
+    },
+    {
+      "epoch": 0.011520508491409277,
+      "grad_norm": 21.142839252245647,
+      "learning_rate": 3.815789473684211e-06,
+      "loss": 0.0233,
+      "step": 58
+    },
+    {
+      "epoch": 0.011719137948157711,
+      "grad_norm": 1.1229291902544376,
+      "learning_rate": 3.8815789473684214e-06,
+      "loss": 0.0183,
+      "step": 59
+    },
+    {
+      "epoch": 0.011917767404906148,
+      "grad_norm": 1.196900422199011,
+      "learning_rate": 3.947368421052632e-06,
+      "loss": 0.0245,
+      "step": 60
+    },
+    {
+      "epoch": 0.012116396861654583,
+      "grad_norm": 1.356686303430086,
+      "learning_rate": 4.013157894736842e-06,
+      "loss": 0.0104,
+      "step": 61
+    },
+    {
+      "epoch": 0.012315026318403019,
+      "grad_norm": 1.0363309431722993,
+      "learning_rate": 4.078947368421053e-06,
+      "loss": 0.0162,
+      "step": 62
+    },
+    {
+      "epoch": 0.012513655775151455,
+      "grad_norm": 1.068841018945078,
+      "learning_rate": 4.144736842105263e-06,
+      "loss": 0.0215,
+      "step": 63
+    },
+    {
+      "epoch": 0.01271228523189989,
+      "grad_norm": 1.3644470124019088,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 0.0308,
+      "step": 64
+    },
+    {
+      "epoch": 0.012910914688648327,
+      "grad_norm": 1.4539722940904474,
+      "learning_rate": 4.276315789473684e-06,
+      "loss": 0.0234,
+      "step": 65
+    },
+    {
+      "epoch": 0.013109544145396763,
+      "grad_norm": 1.8771694709283084,
+      "learning_rate": 4.342105263157895e-06,
+      "loss": 0.0279,
+      "step": 66
+    },
+    {
+      "epoch": 0.013308173602145198,
+      "grad_norm": 1.56691962640963,
+      "learning_rate": 4.407894736842105e-06,
+      "loss": 0.0224,
+      "step": 67
+    },
+    {
+      "epoch": 0.013506803058893634,
+      "grad_norm": 1.0323510479136595,
+      "learning_rate": 4.473684210526316e-06,
+      "loss": 0.0223,
+      "step": 68
+    },
+    {
+      "epoch": 0.013705432515642069,
+      "grad_norm": 2.329638995795258,
+      "learning_rate": 4.539473684210527e-06,
+      "loss": 0.0278,
+      "step": 69
+    },
+    {
+      "epoch": 0.013904061972390505,
+      "grad_norm": 1.1328939658071375,
+      "learning_rate": 4.605263157894737e-06,
+      "loss": 0.0262,
+      "step": 70
+    },
+    {
+      "epoch": 0.014102691429138942,
+      "grad_norm": 1.5789619228369665,
+      "learning_rate": 4.671052631578948e-06,
+      "loss": 0.0226,
+      "step": 71
+    },
+    {
+      "epoch": 0.014301320885887377,
+      "grad_norm": 0.590549283191568,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 0.0222,
+      "step": 72
+    },
+    {
+      "epoch": 0.014499950342635813,
+      "grad_norm": 2.9484330972444797,
+      "learning_rate": 4.802631578947369e-06,
+      "loss": 0.0289,
+      "step": 73
+    },
+    {
+      "epoch": 0.014698579799384248,
+      "grad_norm": 1.7556389255244587,
+      "learning_rate": 4.8684210526315795e-06,
+      "loss": 0.0253,
+      "step": 74
+    },
+    {
+      "epoch": 0.014897209256132684,
+      "grad_norm": 0.8175815824911603,
+      "learning_rate": 4.9342105263157895e-06,
+      "loss": 0.023,
+      "step": 75
+    },
+    {
+      "epoch": 0.01509583871288112,
+      "grad_norm": 1.9543147397030654,
+      "learning_rate": 5e-06,
+      "loss": 0.024,
+      "step": 76
+    },
+    {
+      "epoch": 0.015294468169629555,
+      "grad_norm": 1.411220111931974,
+      "learning_rate": 5.0657894736842104e-06,
+      "loss": 0.0213,
+      "step": 77
+    },
+    {
+      "epoch": 0.015493097626377992,
+      "grad_norm": 2.5787589856867474,
+      "learning_rate": 5.131578947368422e-06,
+      "loss": 0.0194,
+      "step": 78
+    },
+    {
+      "epoch": 0.01569172708312643,
+      "grad_norm": 1.1433613974937111,
+      "learning_rate": 5.197368421052632e-06,
+      "loss": 0.0203,
+      "step": 79
+    },
+    {
+      "epoch": 0.015890356539874865,
+      "grad_norm": 2.943413275721806,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 0.0202,
+      "step": 80
+    },
+    {
+      "epoch": 0.016088985996623298,
+      "grad_norm": 1.2895517559346665,
+      "learning_rate": 5.328947368421054e-06,
+      "loss": 0.0175,
+      "step": 81
+    },
+    {
+      "epoch": 0.016287615453371734,
+      "grad_norm": 1.2902481789966305,
+      "learning_rate": 5.394736842105264e-06,
+      "loss": 0.0273,
+      "step": 82
+    },
+    {
+      "epoch": 0.01648624491012017,
+      "grad_norm": 1.0623933737227271,
+      "learning_rate": 5.460526315789474e-06,
+      "loss": 0.0226,
+      "step": 83
+    },
+    {
+      "epoch": 0.016684874366868607,
+      "grad_norm": 1.551352914024002,
+      "learning_rate": 5.526315789473685e-06,
+      "loss": 0.0227,
+      "step": 84
+    },
+    {
+      "epoch": 0.016883503823617044,
+      "grad_norm": 1.35068564721811,
+      "learning_rate": 5.592105263157896e-06,
+      "loss": 0.0176,
+      "step": 85
+    },
+    {
+      "epoch": 0.017082133280365477,
+      "grad_norm": 0.7178211749767736,
+      "learning_rate": 5.657894736842106e-06,
+      "loss": 0.0161,
+      "step": 86
+    },
+    {
+      "epoch": 0.017280762737113913,
+      "grad_norm": 1.3876820110580605,
+      "learning_rate": 5.723684210526316e-06,
+      "loss": 0.0211,
+      "step": 87
+    },
+    {
+      "epoch": 0.01747939219386235,
+      "grad_norm": 0.9950100462947198,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 0.0236,
+      "step": 88
+    },
+    {
+      "epoch": 0.017678021650610786,
+      "grad_norm": 2.4968422535588806,
+      "learning_rate": 5.855263157894738e-06,
+      "loss": 0.0288,
+      "step": 89
+    },
+    {
+      "epoch": 0.017876651107359223,
+      "grad_norm": 2.6434743230683515,
+      "learning_rate": 5.921052631578948e-06,
+      "loss": 0.03,
+      "step": 90
+    },
+    {
+      "epoch": 0.018075280564107656,
+      "grad_norm": 0.7671799725100394,
+      "learning_rate": 5.9868421052631585e-06,
+      "loss": 0.0303,
+      "step": 91
+    },
+    {
+      "epoch": 0.018273910020856092,
+      "grad_norm": 1.4218642540045263,
+      "learning_rate": 6.0526315789473685e-06,
+      "loss": 0.0265,
+      "step": 92
+    },
+    {
+      "epoch": 0.01847253947760453,
+      "grad_norm": 1.1607669862480212,
+      "learning_rate": 6.118421052631579e-06,
+      "loss": 0.0273,
+      "step": 93
+    },
+    {
+      "epoch": 0.018671168934352965,
+      "grad_norm": 0.5272958683372639,
+      "learning_rate": 6.18421052631579e-06,
+      "loss": 0.0208,
+      "step": 94
+    },
+    {
+      "epoch": 0.0188697983911014,
+      "grad_norm": 0.6384994881210974,
+      "learning_rate": 6.25e-06,
+      "loss": 0.0193,
+      "step": 95
+    },
+    {
+      "epoch": 0.019068427847849838,
+      "grad_norm": 0.4305083797700706,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 0.0155,
+      "step": 96
+    },
+    {
+      "epoch": 0.01926705730459827,
+      "grad_norm": 1.5613150520545582,
+      "learning_rate": 6.381578947368422e-06,
+      "loss": 0.0287,
+      "step": 97
+    },
+    {
+      "epoch": 0.019465686761346707,
+      "grad_norm": 1.207629220791821,
+      "learning_rate": 6.447368421052632e-06,
+      "loss": 0.0126,
+      "step": 98
+    },
+    {
+      "epoch": 0.019664316218095144,
+      "grad_norm": 2.0635349948166946,
+      "learning_rate": 6.513157894736842e-06,
+      "loss": 0.0267,
+      "step": 99
+    },
+    {
+      "epoch": 0.01986294567484358,
+      "grad_norm": 1.725733944785907,
+      "learning_rate": 6.578947368421054e-06,
+      "loss": 0.0307,
+      "step": 100
+    },
+    {
+      "epoch": 0.020061575131592017,
+      "grad_norm": 0.7480249394267019,
+      "learning_rate": 6.644736842105264e-06,
+      "loss": 0.0299,
+      "step": 101
+    },
+    {
+      "epoch": 0.02026020458834045,
+      "grad_norm": 3.894533969012179,
+      "learning_rate": 6.710526315789474e-06,
+      "loss": 0.036,
+      "step": 102
+    },
+    {
+      "epoch": 0.020458834045088886,
+      "grad_norm": 1.9782432860217083,
+      "learning_rate": 6.776315789473686e-06,
+      "loss": 0.019,
+      "step": 103
+    },
+    {
+      "epoch": 0.020657463501837323,
+      "grad_norm": 2.1605252856710577,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 0.0147,
+      "step": 104
+    },
+    {
+      "epoch": 0.02085609295858576,
+      "grad_norm": 1.4740136865902298,
+      "learning_rate": 6.907894736842106e-06,
+      "loss": 0.0225,
+      "step": 105
+    },
+    {
+      "epoch": 0.021054722415334196,
+      "grad_norm": 0.6354162758488423,
+      "learning_rate": 6.973684210526316e-06,
+      "loss": 0.0248,
+      "step": 106
+    },
+    {
+      "epoch": 0.02125335187208263,
+      "grad_norm": 2.7836091692257434,
+      "learning_rate": 7.0394736842105274e-06,
+      "loss": 0.0239,
+      "step": 107
+    },
+    {
+      "epoch": 0.021451981328831065,
+      "grad_norm": 1.1359242167543848,
+      "learning_rate": 7.1052631578947375e-06,
+      "loss": 0.0202,
+      "step": 108
+    },
+    {
+      "epoch": 0.0216506107855795,
+      "grad_norm": 0.9730451625121374,
+      "learning_rate": 7.1710526315789475e-06,
+      "loss": 0.0164,
+      "step": 109
+    },
+    {
+      "epoch": 0.021849240242327938,
+      "grad_norm": 0.73973440231317,
+      "learning_rate": 7.236842105263158e-06,
+      "loss": 0.0208,
+      "step": 110
+    },
+    {
+      "epoch": 0.022047869699076374,
+      "grad_norm": 2.699646318968425,
+      "learning_rate": 7.302631578947369e-06,
+      "loss": 0.0233,
+      "step": 111
+    },
+    {
+      "epoch": 0.022246499155824807,
+      "grad_norm": 1.2218290054724747,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 0.0194,
+      "step": 112
+    },
+    {
+      "epoch": 0.022445128612573244,
+      "grad_norm": 3.2868701187822915,
+      "learning_rate": 7.43421052631579e-06,
+      "loss": 0.0339,
+      "step": 113
+    },
+    {
+      "epoch": 0.02264375806932168,
+      "grad_norm": 2.5641439239722814,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.0286,
+      "step": 114
+    },
+    {
+      "epoch": 0.022842387526070117,
+      "grad_norm": 0.6952134707398119,
+      "learning_rate": 7.565789473684211e-06,
+      "loss": 0.0212,
+      "step": 115
+    },
+    {
+      "epoch": 0.023041016982818553,
+      "grad_norm": 2.011954955233892,
+      "learning_rate": 7.631578947368423e-06,
+      "loss": 0.0259,
+      "step": 116
+    },
+    {
+      "epoch": 0.023239646439566986,
+      "grad_norm": 0.7803532263465452,
+      "learning_rate": 7.697368421052632e-06,
+      "loss": 0.016,
+      "step": 117
+    },
+    {
+      "epoch": 0.023438275896315423,
+      "grad_norm": 1.9847980365877933,
+      "learning_rate": 7.763157894736843e-06,
+      "loss": 0.0195,
+      "step": 118
+    },
+    {
+      "epoch": 0.02363690535306386,
+      "grad_norm": 1.1441961651491355,
+      "learning_rate": 7.828947368421054e-06,
+      "loss": 0.0211,
+      "step": 119
+    },
+    {
+      "epoch": 0.023835534809812296,
+      "grad_norm": 0.6232275208284637,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 0.0155,
+      "step": 120
+    },
+    {
+      "epoch": 0.024034164266560732,
+      "grad_norm": 1.8988678704709767,
+      "learning_rate": 7.960526315789474e-06,
+      "loss": 0.0146,
+      "step": 121
+    },
+    {
+      "epoch": 0.024232793723309165,
+      "grad_norm": 2.105837941919798,
+      "learning_rate": 8.026315789473685e-06,
+      "loss": 0.0295,
+      "step": 122
+    },
+    {
+      "epoch": 0.0244314231800576,
+      "grad_norm": 1.3171076410792975,
+      "learning_rate": 8.092105263157896e-06,
+      "loss": 0.0264,
+      "step": 123
+    },
+    {
+      "epoch": 0.024630052636806038,
+      "grad_norm": 1.5991904078839954,
+      "learning_rate": 8.157894736842106e-06,
+      "loss": 0.0244,
+      "step": 124
+    },
+    {
+      "epoch": 0.024828682093554474,
+      "grad_norm": 1.1618318045886942,
+      "learning_rate": 8.223684210526316e-06,
+      "loss": 0.0201,
+      "step": 125
+    },
+    {
+      "epoch": 0.02502731155030291,
+      "grad_norm": 0.800623192132822,
+      "learning_rate": 8.289473684210526e-06,
+      "loss": 0.0177,
+      "step": 126
+    },
+    {
+      "epoch": 0.025225941007051347,
+      "grad_norm": 1.6211464121453392,
+      "learning_rate": 8.355263157894737e-06,
+      "loss": 0.0155,
+      "step": 127
+    },
+    {
+      "epoch": 0.02542457046379978,
+      "grad_norm": 0.7412717741680597,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 0.0118,
+      "step": 128
+    },
+    {
+      "epoch": 0.025623199920548217,
+      "grad_norm": 0.7011545084516744,
+      "learning_rate": 8.486842105263159e-06,
+      "loss": 0.0197,
+      "step": 129
+    },
+    {
+      "epoch": 0.025821829377296653,
+      "grad_norm": 1.5277548264021221,
+      "learning_rate": 8.552631578947368e-06,
+      "loss": 0.0139,
+      "step": 130
+    },
+    {
+      "epoch": 0.02602045883404509,
+      "grad_norm": 0.8807370373367641,
+      "learning_rate": 8.61842105263158e-06,
+      "loss": 0.0311,
+      "step": 131
+    },
+    {
+      "epoch": 0.026219088290793526,
+      "grad_norm": 1.07819195884304,
+      "learning_rate": 8.68421052631579e-06,
+      "loss": 0.0318,
+      "step": 132
+    },
+    {
+      "epoch": 0.02641771774754196,
+      "grad_norm": 0.696442786177907,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.0234,
+      "step": 133
+    },
+    {
+      "epoch": 0.026616347204290396,
+      "grad_norm": 0.748931405611534,
+      "learning_rate": 8.81578947368421e-06,
+      "loss": 0.0216,
+      "step": 134
+    },
+    {
+      "epoch": 0.026814976661038832,
+      "grad_norm": 0.8573350069478229,
+      "learning_rate": 8.881578947368423e-06,
+      "loss": 0.0291,
+      "step": 135
+    },
+    {
+      "epoch": 0.02701360611778727,
+      "grad_norm": 0.9745147210519538,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 0.0239,
+      "step": 136
+    },
+    {
+      "epoch": 0.027212235574535705,
+      "grad_norm": 0.9802008111815383,
+      "learning_rate": 9.013157894736843e-06,
+      "loss": 0.0209,
+      "step": 137
+    },
+    {
+      "epoch": 0.027410865031284138,
+      "grad_norm": 2.330229398267877,
+      "learning_rate": 9.078947368421054e-06,
+      "loss": 0.0244,
+      "step": 138
+    },
+    {
+      "epoch": 0.027609494488032574,
+      "grad_norm": 1.0656034579377593,
+      "learning_rate": 9.144736842105264e-06,
+      "loss": 0.0159,
+      "step": 139
+    },
+    {
+      "epoch": 0.02780812394478101,
+      "grad_norm": 2.0077938210120796,
+      "learning_rate": 9.210526315789474e-06,
+      "loss": 0.0199,
+      "step": 140
+    },
+    {
+      "epoch": 0.028006753401529447,
+      "grad_norm": 2.635982865291799,
+      "learning_rate": 9.276315789473686e-06,
+      "loss": 0.0276,
+      "step": 141
+    },
+    {
+      "epoch": 0.028205382858277884,
+      "grad_norm": 1.3234128327058614,
+      "learning_rate": 9.342105263157895e-06,
+      "loss": 0.0197,
+      "step": 142
+    },
+    {
+      "epoch": 0.028404012315026317,
+      "grad_norm": 2.8381965627330588,
+      "learning_rate": 9.407894736842106e-06,
+      "loss": 0.0256,
+      "step": 143
+    },
+    {
+      "epoch": 0.028602641771774753,
+      "grad_norm": 2.2353630682958237,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 0.0263,
+      "step": 144
+    },
+    {
+      "epoch": 0.02880127122852319,
+      "grad_norm": 2.609402334542464,
+      "learning_rate": 9.539473684210528e-06,
+      "loss": 0.0208,
+      "step": 145
+    },
+    {
+      "epoch": 0.028999900685271626,
+      "grad_norm": 1.6433963174070783,
+      "learning_rate": 9.605263157894737e-06,
+      "loss": 0.0156,
+      "step": 146
+    },
+    {
+      "epoch": 0.029198530142020063,
+      "grad_norm": 1.2120241439087565,
+      "learning_rate": 9.671052631578948e-06,
+      "loss": 0.0247,
+      "step": 147
+    },
+    {
+      "epoch": 0.029397159598768496,
+      "grad_norm": 2.393880841544898,
+      "learning_rate": 9.736842105263159e-06,
+      "loss": 0.0199,
+      "step": 148
+    },
+    {
+      "epoch": 0.029595789055516932,
+      "grad_norm": 1.147929304198294,
+      "learning_rate": 9.80263157894737e-06,
+      "loss": 0.0142,
+      "step": 149
+    },
+    {
+      "epoch": 0.02979441851226537,
+      "grad_norm": 0.7301143117977119,
+      "learning_rate": 9.868421052631579e-06,
+      "loss": 0.0268,
+      "step": 150
+    },
+    {
+      "epoch": 0.029993047969013805,
+      "grad_norm": 0.7620302081891749,
+      "learning_rate": 9.93421052631579e-06,
+      "loss": 0.0209,
+      "step": 151
+    },
+    {
+      "epoch": 0.03019167742576224,
+      "grad_norm": 1.6325286682577373,
+      "learning_rate": 1e-05,
+      "loss": 0.0165,
+      "step": 152
+    },
+    {
+      "epoch": 0.030390306882510678,
+      "grad_norm": 0.8520608847697756,
+      "learning_rate": 9.999998964752504e-06,
+      "loss": 0.023,
+      "step": 153
+    },
+    {
+      "epoch": 0.03058893633925911,
+      "grad_norm": 0.7134823765486196,
+      "learning_rate": 9.999995859010444e-06,
+      "loss": 0.0213,
+      "step": 154
+    },
+    {
+      "epoch": 0.030787565796007547,
+      "grad_norm": 1.4140312748946944,
+      "learning_rate": 9.999990682775105e-06,
+      "loss": 0.0273,
+      "step": 155
+    },
+    {
+      "epoch": 0.030986195252755984,
+      "grad_norm": 0.9491683828558153,
+      "learning_rate": 9.999983436048632e-06,
+      "loss": 0.0222,
+      "step": 156
+    },
+    {
+      "epoch": 0.03118482470950442,
+      "grad_norm": 1.3782834621848912,
+      "learning_rate": 9.999974118834025e-06,
+      "loss": 0.0162,
+      "step": 157
+    },
+    {
+      "epoch": 0.03138345416625286,
+      "grad_norm": 0.9830166709134077,
+      "learning_rate": 9.999962731135142e-06,
+      "loss": 0.017,
+      "step": 158
+    },
+    {
+      "epoch": 0.03158208362300129,
+      "grad_norm": 2.218321788640668,
+      "learning_rate": 9.999949272956699e-06,
+      "loss": 0.0184,
+      "step": 159
+    },
+    {
+      "epoch": 0.03178071307974973,
+      "grad_norm": 3.6396168680949224,
+      "learning_rate": 9.99993374430427e-06,
+      "loss": 0.033,
+      "step": 160
+    },
+    {
+      "epoch": 0.03197934253649816,
+      "grad_norm": 1.701075218546414,
+      "learning_rate": 9.999916145184286e-06,
+      "loss": 0.0191,
+      "step": 161
+    },
+    {
+      "epoch": 0.032177971993246596,
+      "grad_norm": 1.0244474776048476,
+      "learning_rate": 9.999896475604029e-06,
+      "loss": 0.0164,
+      "step": 162
+    },
+    {
+      "epoch": 0.03237660144999503,
+      "grad_norm": 1.7222754194926513,
+      "learning_rate": 9.999874735571652e-06,
+      "loss": 0.0385,
+      "step": 163
+    },
+    {
+      "epoch": 0.03257523090674347,
+      "grad_norm": 2.5845547766369417,
+      "learning_rate": 9.999850925096153e-06,
+      "loss": 0.0331,
+      "step": 164
+    },
+    {
+      "epoch": 0.032773860363491905,
+      "grad_norm": 1.878569691120879,
+      "learning_rate": 9.999825044187392e-06,
+      "loss": 0.0254,
+      "step": 165
+    },
+    {
+      "epoch": 0.03297248982024034,
+      "grad_norm": 0.9378834395475585,
+      "learning_rate": 9.999797092856089e-06,
+      "loss": 0.0262,
+      "step": 166
+    },
+    {
+      "epoch": 0.03317111927698878,
+      "grad_norm": 0.9618425366151521,
+      "learning_rate": 9.999767071113815e-06,
+      "loss": 0.017,
+      "step": 167
+    },
+    {
+      "epoch": 0.033369748733737215,
+      "grad_norm": 1.7287253946197272,
+      "learning_rate": 9.999734978973006e-06,
+      "loss": 0.0218,
+      "step": 168
+    },
+    {
+      "epoch": 0.03356837819048565,
+      "grad_norm": 0.25580217975967984,
+      "learning_rate": 9.999700816446947e-06,
+      "loss": 0.0104,
+      "step": 169
+    },
+    {
+      "epoch": 0.03376700764723409,
+      "grad_norm": 0.7079216618175301,
+      "learning_rate": 9.999664583549788e-06,
+      "loss": 0.0198,
+      "step": 170
+    },
+    {
+      "epoch": 0.033965637103982524,
+      "grad_norm": 0.5742676426294199,
+      "learning_rate": 9.99962628029653e-06,
+      "loss": 0.0261,
+      "step": 171
+    },
+    {
+      "epoch": 0.03416426656073095,
+      "grad_norm": 0.4760990748227476,
+      "learning_rate": 9.999585906703038e-06,
+      "loss": 0.0223,
+      "step": 172
+    },
+    {
+      "epoch": 0.03436289601747939,
+      "grad_norm": 0.6609769994751641,
+      "learning_rate": 9.999543462786028e-06,
+      "loss": 0.0194,
+      "step": 173
+    },
+    {
+      "epoch": 0.034561525474227826,
+      "grad_norm": 1.0985318270522382,
+      "learning_rate": 9.999498948563076e-06,
+      "loss": 0.0153,
+      "step": 174
+    },
+    {
+      "epoch": 0.03476015493097626,
+      "grad_norm": 0.8636922880587703,
+      "learning_rate": 9.999452364052618e-06,
+      "loss": 0.0243,
+      "step": 175
+    },
+    {
+      "epoch": 0.0349587843877247,
+      "grad_norm": 0.5255774215285303,
+      "learning_rate": 9.99940370927394e-06,
+      "loss": 0.0253,
+      "step": 176
+    },
+    {
+      "epoch": 0.035157413844473136,
+      "grad_norm": 0.9472944515341319,
+      "learning_rate": 9.999352984247196e-06,
+      "loss": 0.0172,
+      "step": 177
+    },
+    {
+      "epoch": 0.03535604330122157,
+      "grad_norm": 1.7710708337619419,
+      "learning_rate": 9.999300188993384e-06,
+      "loss": 0.0209,
+      "step": 178
+    },
+    {
+      "epoch": 0.03555467275797001,
+      "grad_norm": 1.62355032296528,
+      "learning_rate": 9.999245323534372e-06,
+      "loss": 0.0183,
+      "step": 179
+    },
+    {
+      "epoch": 0.035753302214718445,
+      "grad_norm": 1.6780041194910134,
+      "learning_rate": 9.999188387892878e-06,
+      "loss": 0.0184,
+      "step": 180
+    },
+    {
+      "epoch": 0.03595193167146688,
+      "grad_norm": 0.960332672150208,
+      "learning_rate": 9.999129382092478e-06,
+      "loss": 0.0121,
+      "step": 181
+    },
+    {
+      "epoch": 0.03615056112821531,
+      "grad_norm": 0.9539342476315544,
+      "learning_rate": 9.999068306157607e-06,
+      "loss": 0.0134,
+      "step": 182
+    },
+    {
+      "epoch": 0.03634919058496375,
+      "grad_norm": 1.6183931386612114,
+      "learning_rate": 9.999005160113558e-06,
+      "loss": 0.0168,
+      "step": 183
+    },
+    {
+      "epoch": 0.036547820041712184,
+      "grad_norm": 0.9338458758519154,
+      "learning_rate": 9.998939943986476e-06,
+      "loss": 0.022,
+      "step": 184
+    },
+    {
+      "epoch": 0.03674644949846062,
+      "grad_norm": 1.9345051819920607,
+      "learning_rate": 9.998872657803371e-06,
+      "loss": 0.0206,
+      "step": 185
+    },
+    {
+      "epoch": 0.03694507895520906,
+      "grad_norm": 1.1603558183362468,
+      "learning_rate": 9.998803301592105e-06,
+      "loss": 0.014,
+      "step": 186
+    },
+    {
+      "epoch": 0.03714370841195749,
+      "grad_norm": 0.9129479190883742,
+      "learning_rate": 9.998731875381398e-06,
+      "loss": 0.0288,
+      "step": 187
+    },
+    {
+      "epoch": 0.03734233786870593,
+      "grad_norm": 0.9875394679980506,
+      "learning_rate": 9.998658379200826e-06,
+      "loss": 0.0302,
+      "step": 188
+    },
+    {
+      "epoch": 0.037540967325454366,
+      "grad_norm": 1.0324491864294387,
+      "learning_rate": 9.998582813080824e-06,
+      "loss": 0.0157,
+      "step": 189
+    },
+    {
+      "epoch": 0.0377395967822028,
+      "grad_norm": 2.5399121017352098,
+      "learning_rate": 9.998505177052686e-06,
+      "loss": 0.0183,
+      "step": 190
+    },
+    {
+      "epoch": 0.03793822623895124,
+      "grad_norm": 1.0608993845566375,
+      "learning_rate": 9.99842547114856e-06,
+      "loss": 0.0142,
+      "step": 191
+    },
+    {
+      "epoch": 0.038136855695699676,
+      "grad_norm": 0.3873255475411284,
+      "learning_rate": 9.99834369540145e-06,
+      "loss": 0.0179,
+      "step": 192
+    },
+    {
+      "epoch": 0.038335485152448105,
+      "grad_norm": 0.580154672924063,
+      "learning_rate": 9.998259849845224e-06,
+      "loss": 0.0163,
+      "step": 193
+    },
+    {
+      "epoch": 0.03853411460919654,
+      "grad_norm": 0.5232842757349848,
+      "learning_rate": 9.9981739345146e-06,
+      "loss": 0.021,
+      "step": 194
+    },
+    {
+      "epoch": 0.03873274406594498,
+      "grad_norm": 0.5816313182605903,
+      "learning_rate": 9.998085949445154e-06,
+      "loss": 0.0122,
+      "step": 195
+    },
+    {
+      "epoch": 0.038931373522693415,
+      "grad_norm": 2.219159268458382,
+      "learning_rate": 9.99799589467332e-06,
+      "loss": 0.0192,
+      "step": 196
+    },
+    {
+      "epoch": 0.03913000297944185,
+      "grad_norm": 2.0126869985493956,
+      "learning_rate": 9.997903770236393e-06,
+      "loss": 0.0222,
+      "step": 197
+    },
+    {
+      "epoch": 0.03932863243619029,
+      "grad_norm": 0.8427003722810853,
+      "learning_rate": 9.99780957617252e-06,
+      "loss": 0.0155,
+      "step": 198
+    },
+    {
+      "epoch": 0.039527261892938724,
+      "grad_norm": 1.839988273195168,
+      "learning_rate": 9.997713312520703e-06,
+      "loss": 0.0177,
+      "step": 199
+    },
+    {
+      "epoch": 0.03972589134968716,
+      "grad_norm": 2.200388293444912,
+      "learning_rate": 9.99761497932081e-06,
+      "loss": 0.0313,
+      "step": 200
+    },
+    {
+      "epoch": 0.0399245208064356,
+      "grad_norm": 0.9877299401379769,
+      "learning_rate": 9.997514576613561e-06,
+      "loss": 0.0157,
+      "step": 201
+    },
+    {
+      "epoch": 0.04012315026318403,
+      "grad_norm": 1.1088588253099985,
+      "learning_rate": 9.99741210444053e-06,
+      "loss": 0.0198,
+      "step": 202
+    },
+    {
+      "epoch": 0.04032177971993246,
+      "grad_norm": 1.7028868040419975,
+      "learning_rate": 9.997307562844148e-06,
+      "loss": 0.0196,
+      "step": 203
+    },
+    {
+      "epoch": 0.0405204091766809,
+      "grad_norm": 1.2881131253386358,
+      "learning_rate": 9.997200951867711e-06,
+      "loss": 0.022,
+      "step": 204
+    },
+    {
+      "epoch": 0.040719038633429336,
+      "grad_norm": 1.1206602404111856,
+      "learning_rate": 9.997092271555364e-06,
+      "loss": 0.03,
+      "step": 205
+    },
+    {
+      "epoch": 0.04091766809017777,
+      "grad_norm": 0.9055361093963828,
+      "learning_rate": 9.996981521952111e-06,
+      "loss": 0.0166,
+      "step": 206
+    },
+    {
+      "epoch": 0.04111629754692621,
+      "grad_norm": 2.5005202676664315,
+      "learning_rate": 9.996868703103815e-06,
+      "loss": 0.0267,
+      "step": 207
+    },
+    {
+      "epoch": 0.041314927003674645,
+      "grad_norm": 1.6337249581649744,
+      "learning_rate": 9.996753815057191e-06,
+      "loss": 0.0191,
+      "step": 208
+    },
+    {
+      "epoch": 0.04151355646042308,
+      "grad_norm": 0.7101831965292786,
+      "learning_rate": 9.996636857859818e-06,
+      "loss": 0.0314,
+      "step": 209
+    },
+    {
+      "epoch": 0.04171218591717152,
+      "grad_norm": 1.5966574187467166,
+      "learning_rate": 9.996517831560123e-06,
+      "loss": 0.0264,
+      "step": 210
+    },
+    {
+      "epoch": 0.041910815373919955,
+      "grad_norm": 1.8007214636848605,
+      "learning_rate": 9.9963967362074e-06,
+      "loss": 0.0231,
+      "step": 211
+    },
+    {
+      "epoch": 0.04210944483066839,
+      "grad_norm": 3.752484634788327,
+      "learning_rate": 9.996273571851793e-06,
+      "loss": 0.0257,
+      "step": 212
+    },
+    {
+      "epoch": 0.04230807428741682,
+      "grad_norm": 2.618494261051629,
+      "learning_rate": 9.996148338544302e-06,
+      "loss": 0.0238,
+      "step": 213
+    },
+    {
+      "epoch": 0.04250670374416526,
+      "grad_norm": 1.9116871575698093,
+      "learning_rate": 9.996021036336786e-06,
+      "loss": 0.0211,
+      "step": 214
+    },
+    {
+      "epoch": 0.042705333200913694,
+      "grad_norm": 0.5453225145113627,
+      "learning_rate": 9.995891665281965e-06,
+      "loss": 0.02,
+      "step": 215
+    },
+    {
+      "epoch": 0.04290396265766213,
+      "grad_norm": 1.6984059813637056,
+      "learning_rate": 9.995760225433407e-06,
+      "loss": 0.0235,
+      "step": 216
+    },
+    {
+      "epoch": 0.043102592114410566,
+      "grad_norm": 1.4880871765884156,
+      "learning_rate": 9.995626716845541e-06,
+      "loss": 0.0168,
+      "step": 217
+    },
+    {
+      "epoch": 0.043301221571159,
+      "grad_norm": 1.4047456117126256,
+      "learning_rate": 9.995491139573657e-06,
+      "loss": 0.0316,
+      "step": 218
+    },
+    {
+      "epoch": 0.04349985102790744,
+      "grad_norm": 1.2909976897550328,
+      "learning_rate": 9.995353493673892e-06,
+      "loss": 0.0137,
+      "step": 219
+    },
+    {
+      "epoch": 0.043698480484655876,
+      "grad_norm": 0.6501623439351141,
+      "learning_rate": 9.99521377920325e-06,
+      "loss": 0.0161,
+      "step": 220
+    },
+    {
+      "epoch": 0.04389710994140431,
+      "grad_norm": 0.8580190311656244,
+      "learning_rate": 9.995071996219584e-06,
+      "loss": 0.018,
+      "step": 221
+    },
+    {
+      "epoch": 0.04409573939815275,
+      "grad_norm": 1.2921080572385664,
+      "learning_rate": 9.994928144781607e-06,
+      "loss": 0.0129,
+      "step": 222
+    },
+    {
+      "epoch": 0.044294368854901185,
+      "grad_norm": 1.8388869881713987,
+      "learning_rate": 9.994782224948885e-06,
+      "loss": 0.0244,
+      "step": 223
+    },
+    {
+      "epoch": 0.044492998311649615,
+      "grad_norm": 0.8904350028720864,
+      "learning_rate": 9.994634236781845e-06,
+      "loss": 0.0285,
+      "step": 224
+    },
+    {
+      "epoch": 0.04469162776839805,
+      "grad_norm": 1.0351479493656937,
+      "learning_rate": 9.994484180341773e-06,
+      "loss": 0.0129,
+      "step": 225
+    },
+    {
+      "epoch": 0.04489025722514649,
+      "grad_norm": 1.6356872953946395,
+      "learning_rate": 9.994332055690801e-06,
+      "loss": 0.0172,
+      "step": 226
+    },
+    {
+      "epoch": 0.045088886681894924,
+      "grad_norm": 2.4427456672158785,
+      "learning_rate": 9.994177862891927e-06,
+      "loss": 0.0262,
+      "step": 227
+    },
+    {
+      "epoch": 0.04528751613864336,
+      "grad_norm": 1.1833855069108572,
+      "learning_rate": 9.994021602009001e-06,
+      "loss": 0.0212,
+      "step": 228
+    },
+    {
+      "epoch": 0.0454861455953918,
+      "grad_norm": 1.3424482002706264,
+      "learning_rate": 9.99386327310673e-06,
+      "loss": 0.013,
+      "step": 229
+    },
+    {
+      "epoch": 0.045684775052140233,
+      "grad_norm": 1.9959735911469776,
+      "learning_rate": 9.99370287625068e-06,
+      "loss": 0.0136,
+      "step": 230
+    },
+    {
+      "epoch": 0.04588340450888867,
+      "grad_norm": 1.0872172455494484,
+      "learning_rate": 9.99354041150727e-06,
+      "loss": 0.0189,
+      "step": 231
+    },
+    {
+      "epoch": 0.046082033965637106,
+      "grad_norm": 0.8886861060907549,
+      "learning_rate": 9.993375878943775e-06,
+      "loss": 0.0248,
+      "step": 232
+    },
+    {
+      "epoch": 0.04628066342238554,
+      "grad_norm": 1.1111175252979908,
+      "learning_rate": 9.99320927862833e-06,
+      "loss": 0.0207,
+      "step": 233
+    },
+    {
+      "epoch": 0.04647929287913397,
+      "grad_norm": 0.351030054160361,
+      "learning_rate": 9.993040610629923e-06,
+      "loss": 0.0169,
+      "step": 234
+    },
+    {
+      "epoch": 0.04667792233588241,
+      "grad_norm": 1.3626687901401482,
+      "learning_rate": 9.992869875018398e-06,
+      "loss": 0.0176,
+      "step": 235
+    },
+    {
+      "epoch": 0.046876551792630845,
+      "grad_norm": 1.3745225587842778,
+      "learning_rate": 9.992697071864459e-06,
+      "loss": 0.02,
+      "step": 236
+    },
+    {
+      "epoch": 0.04707518124937928,
+      "grad_norm": 0.680922388085837,
+      "learning_rate": 9.992522201239661e-06,
+      "loss": 0.0167,
+      "step": 237
+    },
+    {
+      "epoch": 0.04727381070612772,
+      "grad_norm": 0.6405373508660324,
+      "learning_rate": 9.99234526321642e-06,
+      "loss": 0.0134,
+      "step": 238
+    },
+    {
+      "epoch": 0.047472440162876155,
+      "grad_norm": 1.5525613206844335,
+      "learning_rate": 9.992166257868006e-06,
+      "loss": 0.0244,
+      "step": 239
+    },
+    {
+      "epoch": 0.04767106961962459,
+      "grad_norm": 0.5871522946780782,
+      "learning_rate": 9.991985185268543e-06,
+      "loss": 0.0165,
+      "step": 240
+    },
+    {
+      "epoch": 0.04786969907637303,
+      "grad_norm": 0.8140407938901834,
+      "learning_rate": 9.991802045493013e-06,
+      "loss": 0.0108,
+      "step": 241
+    },
+    {
+      "epoch": 0.048068328533121464,
+      "grad_norm": 1.7987643886031939,
+      "learning_rate": 9.991616838617255e-06,
+      "loss": 0.0247,
+      "step": 242
+    },
+    {
+      "epoch": 0.0482669579898699,
+      "grad_norm": 1.4927221184896913,
+      "learning_rate": 9.991429564717964e-06,
+      "loss": 0.0243,
+      "step": 243
+    },
+    {
+      "epoch": 0.04846558744661833,
+      "grad_norm": 1.1469800583144492,
+      "learning_rate": 9.99124022387269e-06,
+      "loss": 0.0265,
+      "step": 244
+    },
+    {
+      "epoch": 0.04866421690336677,
+      "grad_norm": 0.6033563935180898,
+      "learning_rate": 9.991048816159834e-06,
+      "loss": 0.0211,
+      "step": 245
+    },
+    {
+      "epoch": 0.0488628463601152,
+      "grad_norm": 0.9368709020872853,
+      "learning_rate": 9.990855341658662e-06,
+      "loss": 0.023,
+      "step": 246
+    },
+    {
+      "epoch": 0.04906147581686364,
+      "grad_norm": 0.8399028095373245,
+      "learning_rate": 9.990659800449293e-06,
+      "loss": 0.0123,
+      "step": 247
+    },
+    {
+      "epoch": 0.049260105273612076,
+      "grad_norm": 0.4780920805092138,
+      "learning_rate": 9.990462192612698e-06,
+      "loss": 0.0197,
+      "step": 248
+    },
+    {
+      "epoch": 0.04945873473036051,
+      "grad_norm": 0.7218014814045688,
+      "learning_rate": 9.990262518230706e-06,
+      "loss": 0.0213,
+      "step": 249
+    },
+    {
+      "epoch": 0.04965736418710895,
+      "grad_norm": 1.1716975326112173,
+      "learning_rate": 9.990060777386004e-06,
+      "loss": 0.0206,
+      "step": 250
+    },
+    {
+      "epoch": 0.049855993643857385,
+      "grad_norm": 0.4751820690979132,
+      "learning_rate": 9.98985697016213e-06,
+      "loss": 0.026,
+      "step": 251
+    },
+    {
+      "epoch": 0.05005462310060582,
+      "grad_norm": 0.9003900691985227,
+      "learning_rate": 9.989651096643482e-06,
+      "loss": 0.0128,
+      "step": 252
+    },
+    {
+      "epoch": 0.05025325255735426,
+      "grad_norm": 0.9111928166207327,
+      "learning_rate": 9.98944315691531e-06,
+      "loss": 0.0232,
+      "step": 253
+    },
+    {
+      "epoch": 0.050451882014102695,
+      "grad_norm": 0.5545254300785284,
+      "learning_rate": 9.989233151063726e-06,
+      "loss": 0.0172,
+      "step": 254
+    },
+    {
+      "epoch": 0.050650511470851124,
+      "grad_norm": 0.40224669906766397,
+      "learning_rate": 9.98902107917569e-06,
+      "loss": 0.0233,
+      "step": 255
+    },
+    {
+      "epoch": 0.05084914092759956,
+      "grad_norm": 0.5824400395204213,
+      "learning_rate": 9.98880694133902e-06,
+      "loss": 0.0217,
+      "step": 256
+    },
+    {
+      "epoch": 0.051047770384348,
+      "grad_norm": 1.1709383260744466,
+      "learning_rate": 9.988590737642392e-06,
+      "loss": 0.0135,
+      "step": 257
+    },
+    {
+      "epoch": 0.051246399841096434,
+      "grad_norm": 1.5439449052427077,
+      "learning_rate": 9.988372468175335e-06,
+      "loss": 0.0271,
+      "step": 258
+    },
+    {
+      "epoch": 0.05144502929784487,
+      "grad_norm": 0.8457319931931899,
+      "learning_rate": 9.988152133028234e-06,
+      "loss": 0.024,
+      "step": 259
+    },
+    {
+      "epoch": 0.051643658754593307,
+      "grad_norm": 1.3872763327296354,
+      "learning_rate": 9.98792973229233e-06,
+      "loss": 0.0215,
+      "step": 260
+    },
+    {
+      "epoch": 0.05184228821134174,
+      "grad_norm": 2.1654406629105702,
+      "learning_rate": 9.987705266059721e-06,
+      "loss": 0.0227,
+      "step": 261
+    },
+    {
+      "epoch": 0.05204091766809018,
+      "grad_norm": 1.5458480467902418,
+      "learning_rate": 9.987478734423355e-06,
+      "loss": 0.0218,
+      "step": 262
+    },
+    {
+      "epoch": 0.052239547124838616,
+      "grad_norm": 0.37405484795190086,
+      "learning_rate": 9.987250137477039e-06,
+      "loss": 0.0173,
+      "step": 263
+    },
+    {
+      "epoch": 0.05243817658158705,
+      "grad_norm": 0.7336264315402546,
+      "learning_rate": 9.987019475315437e-06,
+      "loss": 0.018,
+      "step": 264
+    },
+    {
+      "epoch": 0.05263680603833548,
+      "grad_norm": 1.2158389596763395,
+      "learning_rate": 9.986786748034061e-06,
+      "loss": 0.0147,
+      "step": 265
+    },
+    {
+      "epoch": 0.05283543549508392,
+      "grad_norm": 1.130772342954925,
+      "learning_rate": 9.986551955729288e-06,
+      "loss": 0.0263,
+      "step": 266
+    },
+    {
+      "epoch": 0.053034064951832355,
+      "grad_norm": 0.6317791909189016,
+      "learning_rate": 9.986315098498345e-06,
+      "loss": 0.0202,
+      "step": 267
+    },
+    {
+      "epoch": 0.05323269440858079,
+      "grad_norm": 0.44226726309173453,
+      "learning_rate": 9.986076176439313e-06,
+      "loss": 0.0184,
+      "step": 268
+    },
+    {
+      "epoch": 0.05343132386532923,
+      "grad_norm": 0.5474445679262542,
+      "learning_rate": 9.98583518965113e-06,
+      "loss": 0.0095,
+      "step": 269
+    },
+    {
+      "epoch": 0.053629953322077664,
+      "grad_norm": 1.7508003601141047,
+      "learning_rate": 9.985592138233586e-06,
+      "loss": 0.0247,
+      "step": 270
+    },
+    {
+      "epoch": 0.0538285827788261,
+      "grad_norm": 1.5983209020045153,
+      "learning_rate": 9.98534702228733e-06,
+      "loss": 0.0259,
+      "step": 271
+    },
+    {
+      "epoch": 0.05402721223557454,
+      "grad_norm": 1.7251758674302107,
+      "learning_rate": 9.985099841913867e-06,
+      "loss": 0.0107,
+      "step": 272
+    },
+    {
+      "epoch": 0.054225841692322974,
+      "grad_norm": 0.4747190048835185,
+      "learning_rate": 9.984850597215551e-06,
+      "loss": 0.0135,
+      "step": 273
+    },
+    {
+      "epoch": 0.05442447114907141,
+      "grad_norm": 0.9285709587358324,
+      "learning_rate": 9.984599288295593e-06,
+      "loss": 0.017,
+      "step": 274
+    },
+    {
+      "epoch": 0.054623100605819846,
+      "grad_norm": 1.6254232235138268,
+      "learning_rate": 9.984345915258063e-06,
+      "loss": 0.0177,
+      "step": 275
+    },
+    {
+      "epoch": 0.054821730062568276,
+      "grad_norm": 0.8282505450106532,
+      "learning_rate": 9.98409047820788e-06,
+      "loss": 0.0188,
+      "step": 276
+    },
+    {
+      "epoch": 0.05502035951931671,
+      "grad_norm": 1.10441200036311,
+      "learning_rate": 9.983832977250822e-06,
+      "loss": 0.0334,
+      "step": 277
+    },
+    {
+      "epoch": 0.05521898897606515,
+      "grad_norm": 0.4476905849331348,
+      "learning_rate": 9.983573412493519e-06,
+      "loss": 0.0123,
+      "step": 278
+    },
+    {
+      "epoch": 0.055417618432813585,
+      "grad_norm": 0.41407971370056407,
+      "learning_rate": 9.983311784043457e-06,
+      "loss": 0.0163,
+      "step": 279
+    },
+    {
+      "epoch": 0.05561624788956202,
+      "grad_norm": 1.733780241117457,
+      "learning_rate": 9.983048092008973e-06,
+      "loss": 0.0204,
+      "step": 280
+    },
+    {
+      "epoch": 0.05581487734631046,
+      "grad_norm": 0.7561713460915898,
+      "learning_rate": 9.982782336499267e-06,
+      "loss": 0.0191,
+      "step": 281
+    },
+    {
+      "epoch": 0.056013506803058895,
+      "grad_norm": 0.8952250678333094,
+      "learning_rate": 9.982514517624385e-06,
+      "loss": 0.0195,
+      "step": 282
+    },
+    {
+      "epoch": 0.05621213625980733,
+      "grad_norm": 0.33837803654892235,
+      "learning_rate": 9.982244635495232e-06,
+      "loss": 0.0128,
+      "step": 283
+    },
+    {
+      "epoch": 0.05641076571655577,
+      "grad_norm": 0.3491614108577355,
+      "learning_rate": 9.981972690223561e-06,
+      "loss": 0.0104,
+      "step": 284
+    },
+    {
+      "epoch": 0.056609395173304204,
+      "grad_norm": 0.3885685718301126,
+      "learning_rate": 9.98169868192199e-06,
+      "loss": 0.014,
+      "step": 285
+    },
+    {
+      "epoch": 0.056808024630052634,
+      "grad_norm": 1.907227579601552,
+      "learning_rate": 9.981422610703983e-06,
+      "loss": 0.0226,
+      "step": 286
+    },
+    {
+      "epoch": 0.05700665408680107,
+      "grad_norm": 0.6383147141508525,
+      "learning_rate": 9.981144476683863e-06,
+      "loss": 0.0235,
+      "step": 287
+    },
+    {
+      "epoch": 0.05720528354354951,
+      "grad_norm": 1.0623634554315362,
+      "learning_rate": 9.980864279976803e-06,
+      "loss": 0.0199,
+      "step": 288
+    },
+    {
+      "epoch": 0.05740391300029794,
+      "grad_norm": 0.5121142099458341,
+      "learning_rate": 9.980582020698832e-06,
+      "loss": 0.0161,
+      "step": 289
+    },
+    {
+      "epoch": 0.05760254245704638,
+      "grad_norm": 0.8603769100451026,
+      "learning_rate": 9.980297698966835e-06,
+      "loss": 0.0124,
+      "step": 290
+    },
+    {
+      "epoch": 0.057801171913794816,
+      "grad_norm": 1.3969334526557426,
+      "learning_rate": 9.980011314898546e-06,
+      "loss": 0.0175,
+      "step": 291
+    },
+    {
+      "epoch": 0.05799980137054325,
+      "grad_norm": 0.5708269650845633,
+      "learning_rate": 9.97972286861256e-06,
+      "loss": 0.0148,
+      "step": 292
+    },
+    {
+      "epoch": 0.05819843082729169,
+      "grad_norm": 0.6229104672334838,
+      "learning_rate": 9.979432360228322e-06,
+      "loss": 0.0168,
+      "step": 293
+    },
+    {
+      "epoch": 0.058397060284040125,
+      "grad_norm": 0.615731320985735,
+      "learning_rate": 9.979139789866129e-06,
+      "loss": 0.0134,
+      "step": 294
+    },
+    {
+      "epoch": 0.05859568974078856,
+      "grad_norm": 0.625395130998787,
+      "learning_rate": 9.978845157647136e-06,
+      "loss": 0.0097,
+      "step": 295
+    },
+    {
+      "epoch": 0.05879431919753699,
+      "grad_norm": 0.8790404848653711,
+      "learning_rate": 9.978548463693348e-06,
+      "loss": 0.0166,
+      "step": 296
+    },
+    {
+      "epoch": 0.05899294865428543,
+      "grad_norm": 1.4772807736065832,
+      "learning_rate": 9.978249708127627e-06,
+      "loss": 0.0123,
+      "step": 297
+    },
+    {
+      "epoch": 0.059191578111033864,
+      "grad_norm": 0.5627615829863155,
+      "learning_rate": 9.977948891073688e-06,
+      "loss": 0.0151,
+      "step": 298
+    },
+    {
+      "epoch": 0.0593902075677823,
+      "grad_norm": 0.917332306356399,
+      "learning_rate": 9.977646012656099e-06,
+      "loss": 0.0151,
+      "step": 299
+    },
+    {
+      "epoch": 0.05958883702453074,
+      "grad_norm": 0.8670957566639708,
+      "learning_rate": 9.977341073000278e-06,
+      "loss": 0.0167,
+      "step": 300
+    },
+    {
+      "epoch": 0.059787466481279174,
+      "grad_norm": 1.1414866808620485,
+      "learning_rate": 9.977034072232506e-06,
+      "loss": 0.0203,
+      "step": 301
+    },
+    {
+      "epoch": 0.05998609593802761,
+      "grad_norm": 1.1251288241040425,
+      "learning_rate": 9.976725010479907e-06,
+      "loss": 0.0194,
+      "step": 302
+    },
+    {
+      "epoch": 0.06018472539477605,
+      "grad_norm": 0.683122296823212,
+      "learning_rate": 9.976413887870466e-06,
+      "loss": 0.016,
+      "step": 303
+    },
+    {
+      "epoch": 0.06038335485152448,
+      "grad_norm": 0.8996616395705849,
+      "learning_rate": 9.976100704533018e-06,
+      "loss": 0.0223,
+      "step": 304
+    },
+    {
+      "epoch": 0.06058198430827292,
+      "grad_norm": 0.6428548080123838,
+      "learning_rate": 9.97578546059725e-06,
+      "loss": 0.015,
+      "step": 305
+    },
+    {
+      "epoch": 0.060780613765021356,
+      "grad_norm": 1.940990193850758,
+      "learning_rate": 9.975468156193706e-06,
+      "loss": 0.028,
+      "step": 306
+    },
+    {
+      "epoch": 0.060979243221769786,
+      "grad_norm": 1.176442185876376,
+      "learning_rate": 9.975148791453781e-06,
+      "loss": 0.021,
+      "step": 307
+    },
+    {
+      "epoch": 0.06117787267851822,
+      "grad_norm": 0.5811349856371228,
+      "learning_rate": 9.974827366509725e-06,
+      "loss": 0.0164,
+      "step": 308
+    },
+    {
+      "epoch": 0.06137650213526666,
+      "grad_norm": 0.668480148637617,
+      "learning_rate": 9.974503881494638e-06,
+      "loss": 0.0144,
+      "step": 309
+    },
+    {
+      "epoch": 0.061575131592015095,
+      "grad_norm": 1.1370989077273732,
+      "learning_rate": 9.974178336542473e-06,
+      "loss": 0.0163,
+      "step": 310
+    },
+    {
+      "epoch": 0.06177376104876353,
+      "grad_norm": 1.2252608949279702,
+      "learning_rate": 9.973850731788041e-06,
+      "loss": 0.028,
+      "step": 311
+    },
+    {
+      "epoch": 0.06197239050551197,
+      "grad_norm": 1.091045811315918,
+      "learning_rate": 9.973521067367005e-06,
+      "loss": 0.0194,
+      "step": 312
+    },
+    {
+      "epoch": 0.062171019962260404,
+      "grad_norm": 0.5664431447755225,
+      "learning_rate": 9.973189343415872e-06,
+      "loss": 0.0207,
+      "step": 313
+    },
+    {
+      "epoch": 0.06236964941900884,
+      "grad_norm": 0.7809890625347147,
+      "learning_rate": 9.972855560072014e-06,
+      "loss": 0.0186,
+      "step": 314
+    },
+    {
+      "epoch": 0.06256827887575728,
+      "grad_norm": 0.4914092006339943,
+      "learning_rate": 9.972519717473647e-06,
+      "loss": 0.0208,
+      "step": 315
+    },
+    {
+      "epoch": 0.06276690833250571,
+      "grad_norm": 1.352749398041195,
+      "learning_rate": 9.972181815759848e-06,
+      "loss": 0.0178,
+      "step": 316
+    },
+    {
+      "epoch": 0.06296553778925415,
+      "grad_norm": 1.6842469979290473,
+      "learning_rate": 9.971841855070535e-06,
+      "loss": 0.0212,
+      "step": 317
+    },
+    {
+      "epoch": 0.06316416724600259,
+      "grad_norm": 1.181704020131392,
+      "learning_rate": 9.97149983554649e-06,
+      "loss": 0.0249,
+      "step": 318
+    },
+    {
+      "epoch": 0.06336279670275102,
+      "grad_norm": 0.47050532143708484,
+      "learning_rate": 9.971155757329341e-06,
+      "loss": 0.0171,
+      "step": 319
+    },
+    {
+      "epoch": 0.06356142615949946,
+      "grad_norm": 0.8319540289077562,
+      "learning_rate": 9.970809620561573e-06,
+      "loss": 0.0197,
+      "step": 320
+    },
+    {
+      "epoch": 0.0637600556162479,
+      "grad_norm": 2.00049297860715,
+      "learning_rate": 9.970461425386518e-06,
+      "loss": 0.0233,
+      "step": 321
+    },
+    {
+      "epoch": 0.06395868507299632,
+      "grad_norm": 0.6876204488927999,
+      "learning_rate": 9.970111171948362e-06,
+      "loss": 0.0218,
+      "step": 322
+    },
+    {
+      "epoch": 0.06415731452974476,
+      "grad_norm": 1.1508387565108984,
+      "learning_rate": 9.969758860392148e-06,
+      "loss": 0.0254,
+      "step": 323
+    },
+    {
+      "epoch": 0.06435594398649319,
+      "grad_norm": 1.0175823754757551,
+      "learning_rate": 9.96940449086377e-06,
+      "loss": 0.0207,
+      "step": 324
+    },
+    {
+      "epoch": 0.06455457344324163,
+      "grad_norm": 1.342558930748065,
+      "learning_rate": 9.969048063509965e-06,
+      "loss": 0.0166,
+      "step": 325
+    },
+    {
+      "epoch": 0.06475320289999006,
+      "grad_norm": 1.618591294235281,
+      "learning_rate": 9.968689578478334e-06,
+      "loss": 0.017,
+      "step": 326
+    },
+    {
+      "epoch": 0.0649518323567385,
+      "grad_norm": 0.8029753682738998,
+      "learning_rate": 9.968329035917326e-06,
+      "loss": 0.0134,
+      "step": 327
+    },
+    {
+      "epoch": 0.06515046181348694,
+      "grad_norm": 1.4258924678587896,
+      "learning_rate": 9.967966435976237e-06,
+      "loss": 0.0209,
+      "step": 328
+    },
+    {
+      "epoch": 0.06534909127023537,
+      "grad_norm": 0.6264481978575003,
+      "learning_rate": 9.967601778805225e-06,
+      "loss": 0.022,
+      "step": 329
+    },
+    {
+      "epoch": 0.06554772072698381,
+      "grad_norm": 0.9417193977601217,
+      "learning_rate": 9.967235064555289e-06,
+      "loss": 0.0182,
+      "step": 330
+    },
+    {
+      "epoch": 0.06574635018373225,
+      "grad_norm": 1.4600362140153005,
+      "learning_rate": 9.966866293378287e-06,
+      "loss": 0.0169,
+      "step": 331
+    },
+    {
+      "epoch": 0.06594497964048068,
+      "grad_norm": 0.6608946589876135,
+      "learning_rate": 9.966495465426927e-06,
+      "loss": 0.0166,
+      "step": 332
+    },
+    {
+      "epoch": 0.06614360909722912,
+      "grad_norm": 0.5334956998951328,
+      "learning_rate": 9.96612258085477e-06,
+      "loss": 0.0106,
+      "step": 333
+    },
+    {
+      "epoch": 0.06634223855397756,
+      "grad_norm": 0.7085312138032311,
+      "learning_rate": 9.965747639816224e-06,
+      "loss": 0.0147,
+      "step": 334
+    },
+    {
+      "epoch": 0.06654086801072599,
+      "grad_norm": 1.5381074566151718,
+      "learning_rate": 9.965370642466551e-06,
+      "loss": 0.0139,
+      "step": 335
+    },
+    {
+      "epoch": 0.06673949746747443,
+      "grad_norm": 0.6811831504303671,
+      "learning_rate": 9.96499158896187e-06,
+      "loss": 0.0238,
+      "step": 336
+    },
+    {
+      "epoch": 0.06693812692422287,
+      "grad_norm": 1.5626101417205556,
+      "learning_rate": 9.964610479459144e-06,
+      "loss": 0.0229,
+      "step": 337
+    },
+    {
+      "epoch": 0.0671367563809713,
+      "grad_norm": 0.5600383522887019,
+      "learning_rate": 9.964227314116191e-06,
+      "loss": 0.0171,
+      "step": 338
+    },
+    {
+      "epoch": 0.06733538583771974,
+      "grad_norm": 0.8665657897799597,
+      "learning_rate": 9.963842093091678e-06,
+      "loss": 0.0212,
+      "step": 339
+    },
+    {
+      "epoch": 0.06753401529446817,
+      "grad_norm": 0.6444012965528136,
+      "learning_rate": 9.963454816545124e-06,
+      "loss": 0.0162,
+      "step": 340
+    },
+    {
+      "epoch": 0.06773264475121661,
+      "grad_norm": 1.0179480233604894,
+      "learning_rate": 9.963065484636901e-06,
+      "loss": 0.0159,
+      "step": 341
+    },
+    {
+      "epoch": 0.06793127420796505,
+      "grad_norm": 0.33885609792917437,
+      "learning_rate": 9.962674097528232e-06,
+      "loss": 0.0105,
+      "step": 342
+    },
+    {
+      "epoch": 0.06812990366471347,
+      "grad_norm": 0.31682099229115473,
+      "learning_rate": 9.962280655381189e-06,
+      "loss": 0.0139,
+      "step": 343
+    },
+    {
+      "epoch": 0.0683285331214619,
+      "grad_norm": 1.261532302914878,
+      "learning_rate": 9.961885158358696e-06,
+      "loss": 0.013,
+      "step": 344
+    },
+    {
+      "epoch": 0.06852716257821034,
+      "grad_norm": 1.6126011541924115,
+      "learning_rate": 9.961487606624526e-06,
+      "loss": 0.021,
+      "step": 345
+    },
+    {
+      "epoch": 0.06872579203495878,
+      "grad_norm": 0.5677592432568079,
+      "learning_rate": 9.961088000343308e-06,
+      "loss": 0.0182,
+      "step": 346
+    },
+    {
+      "epoch": 0.06892442149170722,
+      "grad_norm": 0.6292493886004704,
+      "learning_rate": 9.960686339680518e-06,
+      "loss": 0.0177,
+      "step": 347
+    },
+    {
+      "epoch": 0.06912305094845565,
+      "grad_norm": 1.9905966048832493,
+      "learning_rate": 9.960282624802482e-06,
+      "loss": 0.0206,
+      "step": 348
+    },
+    {
+      "epoch": 0.06932168040520409,
+      "grad_norm": 0.6186635733129662,
+      "learning_rate": 9.959876855876378e-06,
+      "loss": 0.0162,
+      "step": 349
+    },
+    {
+      "epoch": 0.06952030986195253,
+      "grad_norm": 1.9229044356070553,
+      "learning_rate": 9.959469033070235e-06,
+      "loss": 0.0248,
+      "step": 350
+    },
+    {
+      "epoch": 0.06971893931870096,
+      "grad_norm": 1.3711744907425025,
+      "learning_rate": 9.959059156552931e-06,
+      "loss": 0.0235,
+      "step": 351
+    },
+    {
+      "epoch": 0.0699175687754494,
+      "grad_norm": 0.39607802774515655,
+      "learning_rate": 9.958647226494198e-06,
+      "loss": 0.0132,
+      "step": 352
+    },
+    {
+      "epoch": 0.07011619823219783,
+      "grad_norm": 1.7233078150248982,
+      "learning_rate": 9.958233243064614e-06,
+      "loss": 0.0168,
+      "step": 353
+    },
+    {
+      "epoch": 0.07031482768894627,
+      "grad_norm": 1.6165097055887545,
+      "learning_rate": 9.95781720643561e-06,
+      "loss": 0.022,
+      "step": 354
+    },
+    {
+      "epoch": 0.07051345714569471,
+      "grad_norm": 1.6300475923467623,
+      "learning_rate": 9.957399116779463e-06,
+      "loss": 0.0155,
+      "step": 355
+    },
+    {
+      "epoch": 0.07071208660244314,
+      "grad_norm": 1.3293283054851412,
+      "learning_rate": 9.95697897426931e-06,
+      "loss": 0.0251,
+      "step": 356
+    },
+    {
+      "epoch": 0.07091071605919158,
+      "grad_norm": 0.7311582517038507,
+      "learning_rate": 9.956556779079124e-06,
+      "loss": 0.0201,
+      "step": 357
+    },
+    {
+      "epoch": 0.07110934551594002,
+      "grad_norm": 1.0352583458625484,
+      "learning_rate": 9.956132531383741e-06,
+      "loss": 0.0179,
+      "step": 358
+    },
+    {
+      "epoch": 0.07130797497268845,
+      "grad_norm": 0.6749175905417537,
+      "learning_rate": 9.95570623135884e-06,
+      "loss": 0.0122,
+      "step": 359
+    },
+    {
+      "epoch": 0.07150660442943689,
+      "grad_norm": 0.7936842430096778,
+      "learning_rate": 9.955277879180951e-06,
+      "loss": 0.0156,
+      "step": 360
+    },
+    {
+      "epoch": 0.07170523388618533,
+      "grad_norm": 0.6066214646810274,
+      "learning_rate": 9.954847475027456e-06,
+      "loss": 0.0257,
+      "step": 361
+    },
+    {
+      "epoch": 0.07190386334293376,
+      "grad_norm": 0.5483587007588743,
+      "learning_rate": 9.954415019076581e-06,
+      "loss": 0.0195,
+      "step": 362
+    },
+    {
+      "epoch": 0.0721024927996822,
+      "grad_norm": 1.8751031257546584,
+      "learning_rate": 9.95398051150741e-06,
+      "loss": 0.0263,
+      "step": 363
+    },
+    {
+      "epoch": 0.07230112225643062,
+      "grad_norm": 0.5121194780452556,
+      "learning_rate": 9.95354395249987e-06,
+      "loss": 0.014,
+      "step": 364
+    },
+    {
+      "epoch": 0.07249975171317906,
+      "grad_norm": 0.4817927132351728,
+      "learning_rate": 9.95310534223474e-06,
+      "loss": 0.0122,
+      "step": 365
+    },
+    {
+      "epoch": 0.0726983811699275,
+      "grad_norm": 0.9420406818271349,
+      "learning_rate": 9.952664680893647e-06,
+      "loss": 0.0265,
+      "step": 366
+    },
+    {
+      "epoch": 0.07289701062667593,
+      "grad_norm": 0.700810731765772,
+      "learning_rate": 9.95222196865907e-06,
+      "loss": 0.0224,
+      "step": 367
+    },
+    {
+      "epoch": 0.07309564008342437,
+      "grad_norm": 1.1831981421496827,
+      "learning_rate": 9.951777205714335e-06,
+      "loss": 0.0157,
+      "step": 368
+    },
+    {
+      "epoch": 0.0732942695401728,
+      "grad_norm": 0.844174382418765,
+      "learning_rate": 9.951330392243619e-06,
+      "loss": 0.0208,
+      "step": 369
+    },
+    {
+      "epoch": 0.07349289899692124,
+      "grad_norm": 0.6648604009959039,
+      "learning_rate": 9.950881528431944e-06,
+      "loss": 0.0201,
+      "step": 370
+    },
+    {
+      "epoch": 0.07369152845366968,
+      "grad_norm": 0.9101607178957163,
+      "learning_rate": 9.950430614465187e-06,
+      "loss": 0.0238,
+      "step": 371
+    },
+    {
+      "epoch": 0.07389015791041811,
+      "grad_norm": 1.1877944371561115,
+      "learning_rate": 9.94997765053007e-06,
+      "loss": 0.0229,
+      "step": 372
+    },
+    {
+      "epoch": 0.07408878736716655,
+      "grad_norm": 1.1414413397286445,
+      "learning_rate": 9.949522636814166e-06,
+      "loss": 0.025,
+      "step": 373
+    },
+    {
+      "epoch": 0.07428741682391499,
+      "grad_norm": 1.0250917508528674,
+      "learning_rate": 9.949065573505894e-06,
+      "loss": 0.0223,
+      "step": 374
+    },
+    {
+      "epoch": 0.07448604628066342,
+      "grad_norm": 1.1184758483425372,
+      "learning_rate": 9.948606460794524e-06,
+      "loss": 0.0164,
+      "step": 375
+    },
+    {
+      "epoch": 0.07468467573741186,
+      "grad_norm": 1.191964024912497,
+      "learning_rate": 9.948145298870173e-06,
+      "loss": 0.0218,
+      "step": 376
+    },
+    {
+      "epoch": 0.0748833051941603,
+      "grad_norm": 1.0225207706121757,
+      "learning_rate": 9.94768208792381e-06,
+      "loss": 0.014,
+      "step": 377
+    },
+    {
+      "epoch": 0.07508193465090873,
+      "grad_norm": 0.8774375657656637,
+      "learning_rate": 9.947216828147249e-06,
+      "loss": 0.0187,
+      "step": 378
+    },
+    {
+      "epoch": 0.07528056410765717,
+      "grad_norm": 0.3736459392817161,
+      "learning_rate": 9.946749519733155e-06,
+      "loss": 0.0086,
+      "step": 379
+    },
+    {
+      "epoch": 0.0754791935644056,
+      "grad_norm": 0.8238665843850358,
+      "learning_rate": 9.946280162875036e-06,
+      "loss": 0.0157,
+      "step": 380
+    },
+    {
+      "epoch": 0.07567782302115404,
+      "grad_norm": 0.45936683210469975,
+      "learning_rate": 9.945808757767256e-06,
+      "loss": 0.0099,
+      "step": 381
+    },
+    {
+      "epoch": 0.07587645247790248,
+      "grad_norm": 2.0051899829645343,
+      "learning_rate": 9.945335304605023e-06,
+      "loss": 0.023,
+      "step": 382
+    },
+    {
+      "epoch": 0.07607508193465091,
+      "grad_norm": 0.6435406420149826,
+      "learning_rate": 9.944859803584392e-06,
+      "loss": 0.0114,
+      "step": 383
+    },
+    {
+      "epoch": 0.07627371139139935,
+      "grad_norm": 0.9399241604433789,
+      "learning_rate": 9.94438225490227e-06,
+      "loss": 0.0127,
+      "step": 384
+    },
+    {
+      "epoch": 0.07647234084814777,
+      "grad_norm": 0.6944344878094227,
+      "learning_rate": 9.943902658756405e-06,
+      "loss": 0.0197,
+      "step": 385
+    },
+    {
+      "epoch": 0.07667097030489621,
+      "grad_norm": 0.551483947123314,
+      "learning_rate": 9.9434210153454e-06,
+      "loss": 0.009,
+      "step": 386
+    },
+    {
+      "epoch": 0.07686959976164465,
+      "grad_norm": 1.5968174034926244,
+      "learning_rate": 9.942937324868706e-06,
+      "loss": 0.0233,
+      "step": 387
+    },
+    {
+      "epoch": 0.07706822921839308,
+      "grad_norm": 0.5848570587556811,
+      "learning_rate": 9.942451587526614e-06,
+      "loss": 0.0145,
+      "step": 388
+    },
+    {
+      "epoch": 0.07726685867514152,
+      "grad_norm": 0.9128747168133394,
+      "learning_rate": 9.94196380352027e-06,
+      "loss": 0.0227,
+      "step": 389
+    },
+    {
+      "epoch": 0.07746548813188996,
+      "grad_norm": 1.7264670251438237,
+      "learning_rate": 9.941473973051662e-06,
+      "loss": 0.0295,
+      "step": 390
+    },
+    {
+      "epoch": 0.07766411758863839,
+      "grad_norm": 0.4937037314512115,
+      "learning_rate": 9.94098209632363e-06,
+      "loss": 0.0097,
+      "step": 391
+    },
+    {
+      "epoch": 0.07786274704538683,
+      "grad_norm": 0.775354918247013,
+      "learning_rate": 9.940488173539863e-06,
+      "loss": 0.0211,
+      "step": 392
+    },
+    {
+      "epoch": 0.07806137650213527,
+      "grad_norm": 0.7131682892913337,
+      "learning_rate": 9.93999220490489e-06,
+      "loss": 0.0199,
+      "step": 393
+    },
+    {
+      "epoch": 0.0782600059588837,
+      "grad_norm": 0.6154638513132034,
+      "learning_rate": 9.93949419062409e-06,
+      "loss": 0.0132,
+      "step": 394
+    },
+    {
+      "epoch": 0.07845863541563214,
+      "grad_norm": 1.1169763475776913,
+      "learning_rate": 9.938994130903693e-06,
+      "loss": 0.0217,
+      "step": 395
+    },
+    {
+      "epoch": 0.07865726487238058,
+      "grad_norm": 0.8043725514476527,
+      "learning_rate": 9.938492025950772e-06,
+      "loss": 0.0095,
+      "step": 396
+    },
+    {
+      "epoch": 0.07885589432912901,
+      "grad_norm": 1.3173221094506762,
+      "learning_rate": 9.937987875973249e-06,
+      "loss": 0.0171,
+      "step": 397
+    },
+    {
+      "epoch": 0.07905452378587745,
+      "grad_norm": 0.5812582606522358,
+      "learning_rate": 9.937481681179892e-06,
+      "loss": 0.0171,
+      "step": 398
+    },
+    {
+      "epoch": 0.07925315324262588,
+      "grad_norm": 0.2793393345374467,
+      "learning_rate": 9.936973441780316e-06,
+      "loss": 0.0135,
+      "step": 399
+    },
+    {
+      "epoch": 0.07945178269937432,
+      "grad_norm": 0.34243922533366083,
+      "learning_rate": 9.936463157984981e-06,
+      "loss": 0.014,
+      "step": 400
+    },
+    {
+      "epoch": 0.07965041215612276,
+      "grad_norm": 0.3322421246762658,
+      "learning_rate": 9.935950830005197e-06,
+      "loss": 0.014,
+      "step": 401
+    },
+    {
+      "epoch": 0.0798490416128712,
+      "grad_norm": 0.3204643392343696,
+      "learning_rate": 9.935436458053115e-06,
+      "loss": 0.0117,
+      "step": 402
+    },
+    {
+      "epoch": 0.08004767106961963,
+      "grad_norm": 0.47957220575662524,
+      "learning_rate": 9.934920042341739e-06,
+      "loss": 0.0101,
+      "step": 403
+    },
+    {
+      "epoch": 0.08024630052636807,
+      "grad_norm": 0.5433074802734997,
+      "learning_rate": 9.934401583084916e-06,
+      "loss": 0.0133,
+      "step": 404
+    },
+    {
+      "epoch": 0.08044492998311649,
+      "grad_norm": 0.983630868002418,
+      "learning_rate": 9.933881080497339e-06,
+      "loss": 0.0128,
+      "step": 405
+    },
+    {
+      "epoch": 0.08064355943986493,
+      "grad_norm": 0.8766875713340728,
+      "learning_rate": 9.933358534794547e-06,
+      "loss": 0.0188,
+      "step": 406
+    },
+    {
+      "epoch": 0.08084218889661336,
+      "grad_norm": 0.7878723380271294,
+      "learning_rate": 9.932833946192925e-06,
+      "loss": 0.0144,
+      "step": 407
+    },
+    {
+      "epoch": 0.0810408183533618,
+      "grad_norm": 0.9810561422094143,
+      "learning_rate": 9.932307314909708e-06,
+      "loss": 0.0207,
+      "step": 408
+    },
+    {
+      "epoch": 0.08123944781011024,
+      "grad_norm": 0.7377525283562892,
+      "learning_rate": 9.93177864116297e-06,
+      "loss": 0.0121,
+      "step": 409
+    },
+    {
+      "epoch": 0.08143807726685867,
+      "grad_norm": 0.46490373025726456,
+      "learning_rate": 9.931247925171636e-06,
+      "loss": 0.0069,
+      "step": 410
+    },
+    {
+      "epoch": 0.08163670672360711,
+      "grad_norm": 0.44936367192605653,
+      "learning_rate": 9.930715167155473e-06,
+      "loss": 0.0107,
+      "step": 411
+    },
+    {
+      "epoch": 0.08183533618035554,
+      "grad_norm": 1.0432452279852802,
+      "learning_rate": 9.930180367335098e-06,
+      "loss": 0.0079,
+      "step": 412
+    },
+    {
+      "epoch": 0.08203396563710398,
+      "grad_norm": 1.8118427025754065,
+      "learning_rate": 9.929643525931971e-06,
+      "loss": 0.0209,
+      "step": 413
+    },
+    {
+      "epoch": 0.08223259509385242,
+      "grad_norm": 2.1277356119975,
+      "learning_rate": 9.929104643168396e-06,
+      "loss": 0.0196,
+      "step": 414
+    },
+    {
+      "epoch": 0.08243122455060085,
+      "grad_norm": 1.0672236447535246,
+      "learning_rate": 9.928563719267525e-06,
+      "loss": 0.0251,
+      "step": 415
+    },
+    {
+      "epoch": 0.08262985400734929,
+      "grad_norm": 0.6372995337769211,
+      "learning_rate": 9.92802075445335e-06,
+      "loss": 0.011,
+      "step": 416
+    },
+    {
+      "epoch": 0.08282848346409773,
+      "grad_norm": 2.6818339589471565,
+      "learning_rate": 9.92747574895072e-06,
+      "loss": 0.0234,
+      "step": 417
+    },
+    {
+      "epoch": 0.08302711292084616,
+      "grad_norm": 1.8534077650526866,
+      "learning_rate": 9.926928702985314e-06,
+      "loss": 0.0191,
+      "step": 418
+    },
+    {
+      "epoch": 0.0832257423775946,
+      "grad_norm": 0.6069645126286759,
+      "learning_rate": 9.926379616783667e-06,
+      "loss": 0.0108,
+      "step": 419
+    },
+    {
+      "epoch": 0.08342437183434304,
+      "grad_norm": 1.3113122989349562,
+      "learning_rate": 9.925828490573156e-06,
+      "loss": 0.0207,
+      "step": 420
+    },
+    {
+      "epoch": 0.08362300129109147,
+      "grad_norm": 0.5645438911441669,
+      "learning_rate": 9.925275324581998e-06,
+      "loss": 0.0099,
+      "step": 421
+    },
+    {
+      "epoch": 0.08382163074783991,
+      "grad_norm": 0.5245065449924216,
+      "learning_rate": 9.92472011903926e-06,
+      "loss": 0.0117,
+      "step": 422
+    },
+    {
+      "epoch": 0.08402026020458835,
+      "grad_norm": 0.851720051198573,
+      "learning_rate": 9.924162874174854e-06,
+      "loss": 0.0196,
+      "step": 423
+    },
+    {
+      "epoch": 0.08421888966133678,
+      "grad_norm": 0.6220395812315587,
+      "learning_rate": 9.923603590219531e-06,
+      "loss": 0.0207,
+      "step": 424
+    },
+    {
+      "epoch": 0.08441751911808522,
+      "grad_norm": 0.8302098723285082,
+      "learning_rate": 9.923042267404893e-06,
+      "loss": 0.0144,
+      "step": 425
+    },
+    {
+      "epoch": 0.08461614857483364,
+      "grad_norm": 0.44409974197466384,
+      "learning_rate": 9.922478905963383e-06,
+      "loss": 0.0139,
+      "step": 426
+    },
+    {
+      "epoch": 0.08481477803158208,
+      "grad_norm": 1.3240937379268272,
+      "learning_rate": 9.921913506128286e-06,
+      "loss": 0.0154,
+      "step": 427
+    },
+    {
+      "epoch": 0.08501340748833051,
+      "grad_norm": 0.526570789542041,
+      "learning_rate": 9.921346068133736e-06,
+      "loss": 0.015,
+      "step": 428
+    },
+    {
+      "epoch": 0.08521203694507895,
+      "grad_norm": 0.35386778919766676,
+      "learning_rate": 9.920776592214707e-06,
+      "loss": 0.0123,
+      "step": 429
+    },
+    {
+      "epoch": 0.08541066640182739,
+      "grad_norm": 0.4338611448698274,
+      "learning_rate": 9.92020507860702e-06,
+      "loss": 0.0107,
+      "step": 430
+    },
+    {
+      "epoch": 0.08560929585857582,
+      "grad_norm": 1.170842642025001,
+      "learning_rate": 9.919631527547336e-06,
+      "loss": 0.0151,
+      "step": 431
+    },
+    {
+      "epoch": 0.08580792531532426,
+      "grad_norm": 0.7363695886885039,
+      "learning_rate": 9.919055939273163e-06,
+      "loss": 0.0152,
+      "step": 432
+    },
+    {
+      "epoch": 0.0860065547720727,
+      "grad_norm": 1.019101053836205,
+      "learning_rate": 9.918478314022852e-06,
+      "loss": 0.0232,
+      "step": 433
+    },
+    {
+      "epoch": 0.08620518422882113,
+      "grad_norm": 0.782125020982536,
+      "learning_rate": 9.917898652035595e-06,
+      "loss": 0.0244,
+      "step": 434
+    },
+    {
+      "epoch": 0.08640381368556957,
+      "grad_norm": 0.6310871428581489,
+      "learning_rate": 9.917316953551434e-06,
+      "loss": 0.0225,
+      "step": 435
+    },
+    {
+      "epoch": 0.086602443142318,
+      "grad_norm": 0.43083561752173805,
+      "learning_rate": 9.916733218811246e-06,
+      "loss": 0.0136,
+      "step": 436
+    },
+    {
+      "epoch": 0.08680107259906644,
+      "grad_norm": 0.8104722545750283,
+      "learning_rate": 9.916147448056755e-06,
+      "loss": 0.0144,
+      "step": 437
+    },
+    {
+      "epoch": 0.08699970205581488,
+      "grad_norm": 0.4778684037166286,
+      "learning_rate": 9.91555964153053e-06,
+      "loss": 0.0128,
+      "step": 438
+    },
+    {
+      "epoch": 0.08719833151256332,
+      "grad_norm": 0.8825546091074031,
+      "learning_rate": 9.914969799475978e-06,
+      "loss": 0.0145,
+      "step": 439
+    },
+    {
+      "epoch": 0.08739696096931175,
+      "grad_norm": 0.600585587690455,
+      "learning_rate": 9.914377922137356e-06,
+      "loss": 0.0114,
+      "step": 440
+    },
+    {
+      "epoch": 0.08759559042606019,
+      "grad_norm": 0.9380929174319077,
+      "learning_rate": 9.913784009759757e-06,
+      "loss": 0.0184,
+      "step": 441
+    },
+    {
+      "epoch": 0.08779421988280862,
+      "grad_norm": 0.7171398264697423,
+      "learning_rate": 9.91318806258912e-06,
+      "loss": 0.0187,
+      "step": 442
+    },
+    {
+      "epoch": 0.08799284933955706,
+      "grad_norm": 0.7288666746333805,
+      "learning_rate": 9.912590080872227e-06,
+      "loss": 0.017,
+      "step": 443
+    },
+    {
+      "epoch": 0.0881914787963055,
+      "grad_norm": 0.8521776900166558,
+      "learning_rate": 9.911990064856703e-06,
+      "loss": 0.0223,
+      "step": 444
+    },
+    {
+      "epoch": 0.08839010825305393,
+      "grad_norm": 0.6528440880111782,
+      "learning_rate": 9.91138801479101e-06,
+      "loss": 0.0183,
+      "step": 445
+    },
+    {
+      "epoch": 0.08858873770980237,
+      "grad_norm": 0.4836930626793002,
+      "learning_rate": 9.910783930924458e-06,
+      "loss": 0.0098,
+      "step": 446
+    },
+    {
+      "epoch": 0.08878736716655079,
+      "grad_norm": 0.998927006114887,
+      "learning_rate": 9.910177813507198e-06,
+      "loss": 0.014,
+      "step": 447
+    },
+    {
+      "epoch": 0.08898599662329923,
+      "grad_norm": 0.8181775259423208,
+      "learning_rate": 9.909569662790224e-06,
+      "loss": 0.0192,
+      "step": 448
+    },
+    {
+      "epoch": 0.08918462608004767,
+      "grad_norm": 0.5583401583915387,
+      "learning_rate": 9.908959479025369e-06,
+      "loss": 0.0119,
+      "step": 449
+    },
+    {
+      "epoch": 0.0893832555367961,
+      "grad_norm": 0.8838426600045396,
+      "learning_rate": 9.908347262465308e-06,
+      "loss": 0.0213,
+      "step": 450
+    },
+    {
+      "epoch": 0.08958188499354454,
+      "grad_norm": 0.7498887472731395,
+      "learning_rate": 9.907733013363563e-06,
+      "loss": 0.0142,
+      "step": 451
+    },
+    {
+      "epoch": 0.08978051445029298,
+      "grad_norm": 0.8966575656361611,
+      "learning_rate": 9.907116731974491e-06,
+      "loss": 0.0193,
+      "step": 452
+    },
+    {
+      "epoch": 0.08997914390704141,
+      "grad_norm": 0.5073403960586594,
+      "learning_rate": 9.906498418553295e-06,
+      "loss": 0.0137,
+      "step": 453
+    },
+    {
+      "epoch": 0.09017777336378985,
+      "grad_norm": 1.1137939580467742,
+      "learning_rate": 9.905878073356015e-06,
+      "loss": 0.0192,
+      "step": 454
+    },
+    {
+      "epoch": 0.09037640282053828,
+      "grad_norm": 1.1344716602210057,
+      "learning_rate": 9.90525569663954e-06,
+      "loss": 0.0161,
+      "step": 455
+    },
+    {
+      "epoch": 0.09057503227728672,
+      "grad_norm": 0.7220695206586469,
+      "learning_rate": 9.904631288661593e-06,
+      "loss": 0.0182,
+      "step": 456
+    },
+    {
+      "epoch": 0.09077366173403516,
+      "grad_norm": 0.5923800565042013,
+      "learning_rate": 9.904004849680741e-06,
+      "loss": 0.0174,
+      "step": 457
+    },
+    {
+      "epoch": 0.0909722911907836,
+      "grad_norm": 1.0827911084412167,
+      "learning_rate": 9.903376379956392e-06,
+      "loss": 0.0129,
+      "step": 458
+    },
+    {
+      "epoch": 0.09117092064753203,
+      "grad_norm": 0.5158773883497091,
+      "learning_rate": 9.902745879748793e-06,
+      "loss": 0.0159,
+      "step": 459
+    },
+    {
+      "epoch": 0.09136955010428047,
+      "grad_norm": 1.0459539619692417,
+      "learning_rate": 9.902113349319035e-06,
+      "loss": 0.0253,
+      "step": 460
+    },
+    {
+      "epoch": 0.0915681795610289,
+      "grad_norm": 1.2859832032862175,
+      "learning_rate": 9.90147878892905e-06,
+      "loss": 0.0192,
+      "step": 461
+    },
+    {
+      "epoch": 0.09176680901777734,
+      "grad_norm": 0.5360090137854244,
+      "learning_rate": 9.900842198841606e-06,
+      "loss": 0.0153,
+      "step": 462
+    },
+    {
+      "epoch": 0.09196543847452578,
+      "grad_norm": 0.4013944962969106,
+      "learning_rate": 9.900203579320316e-06,
+      "loss": 0.0154,
+      "step": 463
+    },
+    {
+      "epoch": 0.09216406793127421,
+      "grad_norm": 0.41838257115387206,
+      "learning_rate": 9.89956293062963e-06,
+      "loss": 0.0128,
+      "step": 464
+    },
+    {
+      "epoch": 0.09236269738802265,
+      "grad_norm": 0.713293197602063,
+      "learning_rate": 9.898920253034841e-06,
+      "loss": 0.0148,
+      "step": 465
+    },
+    {
+      "epoch": 0.09256132684477109,
+      "grad_norm": 0.5124949670835264,
+      "learning_rate": 9.898275546802082e-06,
+      "loss": 0.0182,
+      "step": 466
+    },
+    {
+      "epoch": 0.09275995630151952,
+      "grad_norm": 0.7347929074207598,
+      "learning_rate": 9.897628812198324e-06,
+      "loss": 0.018,
+      "step": 467
+    },
+    {
+      "epoch": 0.09295858575826794,
+      "grad_norm": 0.3702075450464756,
+      "learning_rate": 9.896980049491378e-06,
+      "loss": 0.0153,
+      "step": 468
+    },
+    {
+      "epoch": 0.09315721521501638,
+      "grad_norm": 0.8668215363272558,
+      "learning_rate": 9.896329258949898e-06,
+      "loss": 0.0186,
+      "step": 469
+    },
+    {
+      "epoch": 0.09335584467176482,
+      "grad_norm": 0.9267636519344175,
+      "learning_rate": 9.895676440843376e-06,
+      "loss": 0.02,
+      "step": 470
+    },
+    {
+      "epoch": 0.09355447412851325,
+      "grad_norm": 1.975894079341777,
+      "learning_rate": 9.895021595442143e-06,
+      "loss": 0.0153,
+      "step": 471
+    },
+    {
+      "epoch": 0.09375310358526169,
+      "grad_norm": 0.9664233797506984,
+      "learning_rate": 9.894364723017369e-06,
+      "loss": 0.0248,
+      "step": 472
+    },
+    {
+      "epoch": 0.09395173304201013,
+      "grad_norm": 0.7298025816829822,
+      "learning_rate": 9.893705823841063e-06,
+      "loss": 0.0136,
+      "step": 473
+    },
+    {
+      "epoch": 0.09415036249875856,
+      "grad_norm": 0.5028928969765695,
+      "learning_rate": 9.893044898186077e-06,
+      "loss": 0.0132,
+      "step": 474
+    },
+    {
+      "epoch": 0.094348991955507,
+      "grad_norm": 0.5343396584732334,
+      "learning_rate": 9.8923819463261e-06,
+      "loss": 0.0114,
+      "step": 475
+    },
+    {
+      "epoch": 0.09454762141225544,
+      "grad_norm": 1.043121510446111,
+      "learning_rate": 9.891716968535655e-06,
+      "loss": 0.0233,
+      "step": 476
+    },
+    {
+      "epoch": 0.09474625086900387,
+      "grad_norm": 0.9607158804060207,
+      "learning_rate": 9.891049965090116e-06,
+      "loss": 0.0184,
+      "step": 477
+    },
+    {
+      "epoch": 0.09494488032575231,
+      "grad_norm": 2.04441514855444,
+      "learning_rate": 9.890380936265683e-06,
+      "loss": 0.0261,
+      "step": 478
+    },
+    {
+      "epoch": 0.09514350978250075,
+      "grad_norm": 0.44891259364731184,
+      "learning_rate": 9.8897098823394e-06,
+      "loss": 0.0161,
+      "step": 479
+    },
+    {
+      "epoch": 0.09534213923924918,
+      "grad_norm": 0.20556262152171734,
+      "learning_rate": 9.889036803589154e-06,
+      "loss": 0.0066,
+      "step": 480
+    },
+    {
+      "epoch": 0.09554076869599762,
+      "grad_norm": 0.62530012349153,
+      "learning_rate": 9.888361700293663e-06,
+      "loss": 0.013,
+      "step": 481
+    },
+    {
+      "epoch": 0.09573939815274606,
+      "grad_norm": 0.5060323442897047,
+      "learning_rate": 9.887684572732486e-06,
+      "loss": 0.0176,
+      "step": 482
+    },
+    {
+      "epoch": 0.09593802760949449,
+      "grad_norm": 0.6697767814686015,
+      "learning_rate": 9.887005421186022e-06,
+      "loss": 0.0128,
+      "step": 483
+    },
+    {
+      "epoch": 0.09613665706624293,
+      "grad_norm": 0.49988533088265713,
+      "learning_rate": 9.886324245935508e-06,
+      "loss": 0.0123,
+      "step": 484
+    },
+    {
+      "epoch": 0.09633528652299136,
+      "grad_norm": 0.5219089565515956,
+      "learning_rate": 9.885641047263017e-06,
+      "loss": 0.0245,
+      "step": 485
+    },
+    {
+      "epoch": 0.0965339159797398,
+      "grad_norm": 0.566826196384926,
+      "learning_rate": 9.88495582545146e-06,
+      "loss": 0.0145,
+      "step": 486
+    },
+    {
+      "epoch": 0.09673254543648824,
+      "grad_norm": 0.959759790672288,
+      "learning_rate": 9.88426858078459e-06,
+      "loss": 0.0194,
+      "step": 487
+    },
+    {
+      "epoch": 0.09693117489323666,
+      "grad_norm": 0.9036100999985963,
+      "learning_rate": 9.88357931354699e-06,
+      "loss": 0.0129,
+      "step": 488
+    },
+    {
+      "epoch": 0.0971298043499851,
+      "grad_norm": 0.5669073431377971,
+      "learning_rate": 9.882888024024086e-06,
+      "loss": 0.0093,
+      "step": 489
+    },
+    {
+      "epoch": 0.09732843380673353,
+      "grad_norm": 1.0332312565203494,
+      "learning_rate": 9.882194712502142e-06,
+      "loss": 0.0263,
+      "step": 490
+    },
+    {
+      "epoch": 0.09752706326348197,
+      "grad_norm": 1.4518945147996904,
+      "learning_rate": 9.881499379268258e-06,
+      "loss": 0.0165,
+      "step": 491
+    },
+    {
+      "epoch": 0.0977256927202304,
+      "grad_norm": 0.7165588378220921,
+      "learning_rate": 9.880802024610367e-06,
+      "loss": 0.0261,
+      "step": 492
+    },
+    {
+      "epoch": 0.09792432217697884,
+      "grad_norm": 0.33507821178982644,
+      "learning_rate": 9.880102648817249e-06,
+      "loss": 0.018,
+      "step": 493
+    },
+    {
+      "epoch": 0.09812295163372728,
+      "grad_norm": 1.0953941169607475,
+      "learning_rate": 9.879401252178508e-06,
+      "loss": 0.0163,
+      "step": 494
+    },
+    {
+      "epoch": 0.09832158109047572,
+      "grad_norm": 1.1884751994563965,
+      "learning_rate": 9.878697834984596e-06,
+      "loss": 0.017,
+      "step": 495
+    },
+    {
+      "epoch": 0.09852021054722415,
+      "grad_norm": 1.3895562605165783,
+      "learning_rate": 9.877992397526795e-06,
+      "loss": 0.0178,
+      "step": 496
+    },
+    {
+      "epoch": 0.09871884000397259,
+      "grad_norm": 1.0077625272176394,
+      "learning_rate": 9.877284940097229e-06,
+      "loss": 0.014,
+      "step": 497
+    },
+    {
+      "epoch": 0.09891746946072102,
+      "grad_norm": 1.1176332665051876,
+      "learning_rate": 9.876575462988852e-06,
+      "loss": 0.0176,
+      "step": 498
+    },
+    {
+      "epoch": 0.09911609891746946,
+      "grad_norm": 0.9197946583172847,
+      "learning_rate": 9.87586396649546e-06,
+      "loss": 0.0186,
+      "step": 499
+    },
+    {
+      "epoch": 0.0993147283742179,
+      "grad_norm": 0.664357412818359,
+      "learning_rate": 9.875150450911682e-06,
+      "loss": 0.0078,
+      "step": 500
+    },
+    {
+      "epoch": 0.09951335783096633,
+      "grad_norm": 0.9879849373646336,
+      "learning_rate": 9.874434916532984e-06,
+      "loss": 0.0174,
+      "step": 501
+    },
+    {
+      "epoch": 0.09971198728771477,
+      "grad_norm": 0.8015238994270766,
+      "learning_rate": 9.87371736365567e-06,
+      "loss": 0.0182,
+      "step": 502
+    },
+    {
+      "epoch": 0.09991061674446321,
+      "grad_norm": 2.5006213146117324,
+      "learning_rate": 9.872997792576876e-06,
+      "loss": 0.0192,
+      "step": 503
+    },
+    {
+      "epoch": 0.10010924620121164,
+      "grad_norm": 2.0986832355357263,
+      "learning_rate": 9.872276203594575e-06,
+      "loss": 0.015,
+      "step": 504
+    },
+    {
+      "epoch": 0.10030787565796008,
+      "grad_norm": 1.7712689606880594,
+      "learning_rate": 9.871552597007577e-06,
+      "loss": 0.0162,
+      "step": 505
+    },
+    {
+      "epoch": 0.10050650511470852,
+      "grad_norm": 0.6358423775925522,
+      "learning_rate": 9.870826973115528e-06,
+      "loss": 0.0088,
+      "step": 506
+    },
+    {
+      "epoch": 0.10070513457145695,
+      "grad_norm": 1.2361002395299776,
+      "learning_rate": 9.870099332218908e-06,
+      "loss": 0.0132,
+      "step": 507
+    },
+    {
+      "epoch": 0.10090376402820539,
+      "grad_norm": 0.3894121263307642,
+      "learning_rate": 9.869369674619031e-06,
+      "loss": 0.009,
+      "step": 508
+    },
+    {
+      "epoch": 0.10110239348495381,
+      "grad_norm": 0.6814889911507336,
+      "learning_rate": 9.868638000618047e-06,
+      "loss": 0.018,
+      "step": 509
+    },
+    {
+      "epoch": 0.10130102294170225,
+      "grad_norm": 1.5433233594496414,
+      "learning_rate": 9.867904310518943e-06,
+      "loss": 0.0169,
+      "step": 510
+    },
+    {
+      "epoch": 0.10149965239845068,
+      "grad_norm": 2.086256409582492,
+      "learning_rate": 9.867168604625538e-06,
+      "loss": 0.0272,
+      "step": 511
+    },
+    {
+      "epoch": 0.10169828185519912,
+      "grad_norm": 1.1008955506156466,
+      "learning_rate": 9.86643088324249e-06,
+      "loss": 0.0197,
+      "step": 512
+    },
+    {
+      "epoch": 0.10189691131194756,
+      "grad_norm": 0.6668853742795092,
+      "learning_rate": 9.865691146675286e-06,
+      "loss": 0.014,
+      "step": 513
+    },
+    {
+      "epoch": 0.102095540768696,
+      "grad_norm": 0.6058575326394435,
+      "learning_rate": 9.86494939523025e-06,
+      "loss": 0.0107,
+      "step": 514
+    },
+    {
+      "epoch": 0.10229417022544443,
+      "grad_norm": 0.9452200105677793,
+      "learning_rate": 9.864205629214542e-06,
+      "loss": 0.0224,
+      "step": 515
+    },
+    {
+      "epoch": 0.10249279968219287,
+      "grad_norm": 1.0087103985746844,
+      "learning_rate": 9.863459848936155e-06,
+      "loss": 0.0175,
+      "step": 516
+    },
+    {
+      "epoch": 0.1026914291389413,
+      "grad_norm": 1.4576341787329083,
+      "learning_rate": 9.862712054703913e-06,
+      "loss": 0.017,
+      "step": 517
+    },
+    {
+      "epoch": 0.10289005859568974,
+      "grad_norm": 1.007654984553146,
+      "learning_rate": 9.861962246827479e-06,
+      "loss": 0.0197,
+      "step": 518
+    },
+    {
+      "epoch": 0.10308868805243818,
+      "grad_norm": 1.1768748633973105,
+      "learning_rate": 9.861210425617348e-06,
+      "loss": 0.0146,
+      "step": 519
+    },
+    {
+      "epoch": 0.10328731750918661,
+      "grad_norm": 0.5946983343781779,
+      "learning_rate": 9.860456591384849e-06,
+      "loss": 0.0161,
+      "step": 520
+    },
+    {
+      "epoch": 0.10348594696593505,
+      "grad_norm": 0.8631523397738572,
+      "learning_rate": 9.85970074444214e-06,
+      "loss": 0.0169,
+      "step": 521
+    },
+    {
+      "epoch": 0.10368457642268349,
+      "grad_norm": 0.6524946202638839,
+      "learning_rate": 9.858942885102221e-06,
+      "loss": 0.0231,
+      "step": 522
+    },
+    {
+      "epoch": 0.10388320587943192,
+      "grad_norm": 0.5099019794008004,
+      "learning_rate": 9.85818301367892e-06,
+      "loss": 0.0181,
+      "step": 523
+    },
+    {
+      "epoch": 0.10408183533618036,
+      "grad_norm": 0.503120599970783,
+      "learning_rate": 9.857421130486897e-06,
+      "loss": 0.0172,
+      "step": 524
+    },
+    {
+      "epoch": 0.1042804647929288,
+      "grad_norm": 0.5964107917213637,
+      "learning_rate": 9.85665723584165e-06,
+      "loss": 0.0175,
+      "step": 525
+    },
+    {
+      "epoch": 0.10447909424967723,
+      "grad_norm": 0.6031856884274558,
+      "learning_rate": 9.855891330059502e-06,
+      "loss": 0.0174,
+      "step": 526
+    },
+    {
+      "epoch": 0.10467772370642567,
+      "grad_norm": 0.47774054717960784,
+      "learning_rate": 9.85512341345762e-06,
+      "loss": 0.02,
+      "step": 527
+    },
+    {
+      "epoch": 0.1048763531631741,
+      "grad_norm": 0.47747803803414446,
+      "learning_rate": 9.854353486353994e-06,
+      "loss": 0.0176,
+      "step": 528
+    },
+    {
+      "epoch": 0.10507498261992254,
+      "grad_norm": 0.4118060996615957,
+      "learning_rate": 9.853581549067449e-06,
+      "loss": 0.011,
+      "step": 529
+    },
+    {
+      "epoch": 0.10527361207667096,
+      "grad_norm": 1.7006846993605238,
+      "learning_rate": 9.852807601917647e-06,
+      "loss": 0.0245,
+      "step": 530
+    },
+    {
+      "epoch": 0.1054722415334194,
+      "grad_norm": 1.319886511097232,
+      "learning_rate": 9.852031645225075e-06,
+      "loss": 0.0133,
+      "step": 531
+    },
+    {
+      "epoch": 0.10567087099016784,
+      "grad_norm": 0.5419706523360872,
+      "learning_rate": 9.851253679311059e-06,
+      "loss": 0.0115,
+      "step": 532
+    },
+    {
+      "epoch": 0.10586950044691627,
+      "grad_norm": 0.4436521767303458,
+      "learning_rate": 9.850473704497752e-06,
+      "loss": 0.0173,
+      "step": 533
+    },
+    {
+      "epoch": 0.10606812990366471,
+      "grad_norm": 0.9414894737976166,
+      "learning_rate": 9.84969172110814e-06,
+      "loss": 0.0125,
+      "step": 534
+    },
+    {
+      "epoch": 0.10626675936041315,
+      "grad_norm": 0.5777200514403674,
+      "learning_rate": 9.848907729466045e-06,
+      "loss": 0.0157,
+      "step": 535
+    },
+    {
+      "epoch": 0.10646538881716158,
+      "grad_norm": 0.5350127642077067,
+      "learning_rate": 9.848121729896114e-06,
+      "loss": 0.0109,
+      "step": 536
+    },
+    {
+      "epoch": 0.10666401827391002,
+      "grad_norm": 0.6968674417806208,
+      "learning_rate": 9.84733372272383e-06,
+      "loss": 0.0238,
+      "step": 537
+    },
+    {
+      "epoch": 0.10686264773065846,
+      "grad_norm": 0.6180948340768215,
+      "learning_rate": 9.846543708275507e-06,
+      "loss": 0.0226,
+      "step": 538
+    },
+    {
+      "epoch": 0.10706127718740689,
+      "grad_norm": 0.7782144147736434,
+      "learning_rate": 9.845751686878286e-06,
+      "loss": 0.021,
+      "step": 539
+    },
+    {
+      "epoch": 0.10725990664415533,
+      "grad_norm": 0.8480072268503545,
+      "learning_rate": 9.844957658860143e-06,
+      "loss": 0.0189,
+      "step": 540
+    },
+    {
+      "epoch": 0.10745853610090376,
+      "grad_norm": 0.48593638963667746,
+      "learning_rate": 9.844161624549889e-06,
+      "loss": 0.0148,
+      "step": 541
+    },
+    {
+      "epoch": 0.1076571655576522,
+      "grad_norm": 1.0293729094433886,
+      "learning_rate": 9.843363584277154e-06,
+      "loss": 0.0148,
+      "step": 542
+    },
+    {
+      "epoch": 0.10785579501440064,
+      "grad_norm": 0.5138840542732326,
+      "learning_rate": 9.84256353837241e-06,
+      "loss": 0.0199,
+      "step": 543
+    },
+    {
+      "epoch": 0.10805442447114907,
+      "grad_norm": 0.9923481664493533,
+      "learning_rate": 9.841761487166955e-06,
+      "loss": 0.0209,
+      "step": 544
+    },
+    {
+      "epoch": 0.10825305392789751,
+      "grad_norm": 0.499458338835493,
+      "learning_rate": 9.840957430992917e-06,
+      "loss": 0.0195,
+      "step": 545
+    },
+    {
+      "epoch": 0.10845168338464595,
+      "grad_norm": 0.40579211347456434,
+      "learning_rate": 9.840151370183253e-06,
+      "loss": 0.0153,
+      "step": 546
+    },
+    {
+      "epoch": 0.10865031284139438,
+      "grad_norm": 0.5908120798309358,
+      "learning_rate": 9.839343305071755e-06,
+      "loss": 0.0121,
+      "step": 547
+    },
+    {
+      "epoch": 0.10884894229814282,
+      "grad_norm": 0.8968327120573901,
+      "learning_rate": 9.838533235993041e-06,
+      "loss": 0.0238,
+      "step": 548
+    },
+    {
+      "epoch": 0.10904757175489126,
+      "grad_norm": 1.01470974568391,
+      "learning_rate": 9.837721163282558e-06,
+      "loss": 0.0168,
+      "step": 549
+    },
+    {
+      "epoch": 0.10924620121163969,
+      "grad_norm": 0.3781128929172408,
+      "learning_rate": 9.836907087276587e-06,
+      "loss": 0.0188,
+      "step": 550
+    },
+    {
+      "epoch": 0.10944483066838812,
+      "grad_norm": 0.3520969782131752,
+      "learning_rate": 9.836091008312233e-06,
+      "loss": 0.0144,
+      "step": 551
+    },
+    {
+      "epoch": 0.10964346012513655,
+      "grad_norm": 1.206009624884685,
+      "learning_rate": 9.835272926727439e-06,
+      "loss": 0.0169,
+      "step": 552
+    },
+    {
+      "epoch": 0.10984208958188499,
+      "grad_norm": 0.5829489199164067,
+      "learning_rate": 9.834452842860967e-06,
+      "loss": 0.0141,
+      "step": 553
+    },
+    {
+      "epoch": 0.11004071903863342,
+      "grad_norm": 0.7985221379687066,
+      "learning_rate": 9.833630757052413e-06,
+      "loss": 0.0176,
+      "step": 554
+    },
+    {
+      "epoch": 0.11023934849538186,
+      "grad_norm": 0.7533308379120465,
+      "learning_rate": 9.832806669642203e-06,
+      "loss": 0.0151,
+      "step": 555
+    },
+    {
+      "epoch": 0.1104379779521303,
+      "grad_norm": 1.08877862811918,
+      "learning_rate": 9.831980580971591e-06,
+      "loss": 0.0196,
+      "step": 556
+    },
+    {
+      "epoch": 0.11063660740887873,
+      "grad_norm": 0.8723157647635501,
+      "learning_rate": 9.831152491382658e-06,
+      "loss": 0.0138,
+      "step": 557
+    },
+    {
+      "epoch": 0.11083523686562717,
+      "grad_norm": 1.2697238695545063,
+      "learning_rate": 9.83032240121832e-06,
+      "loss": 0.0225,
+      "step": 558
+    },
+    {
+      "epoch": 0.11103386632237561,
+      "grad_norm": 0.4616713641640445,
+      "learning_rate": 9.82949031082231e-06,
+      "loss": 0.0156,
+      "step": 559
+    },
+    {
+      "epoch": 0.11123249577912404,
+      "grad_norm": 0.349449094735304,
+      "learning_rate": 9.8286562205392e-06,
+      "loss": 0.012,
+      "step": 560
+    },
+    {
+      "epoch": 0.11143112523587248,
+      "grad_norm": 1.1595881223097364,
+      "learning_rate": 9.827820130714383e-06,
+      "loss": 0.0177,
+      "step": 561
+    },
+    {
+      "epoch": 0.11162975469262092,
+      "grad_norm": 0.5231453160678909,
+      "learning_rate": 9.826982041694086e-06,
+      "loss": 0.0086,
+      "step": 562
+    },
+    {
+      "epoch": 0.11182838414936935,
+      "grad_norm": 0.7972550695997788,
+      "learning_rate": 9.826141953825358e-06,
+      "loss": 0.0154,
+      "step": 563
+    },
+    {
+      "epoch": 0.11202701360611779,
+      "grad_norm": 0.4601783531819776,
+      "learning_rate": 9.825299867456082e-06,
+      "loss": 0.0186,
+      "step": 564
+    },
+    {
+      "epoch": 0.11222564306286623,
+      "grad_norm": 0.7214422323486367,
+      "learning_rate": 9.824455782934961e-06,
+      "loss": 0.011,
+      "step": 565
+    },
+    {
+      "epoch": 0.11242427251961466,
+      "grad_norm": 1.680930925926308,
+      "learning_rate": 9.823609700611534e-06,
+      "loss": 0.0181,
+      "step": 566
+    },
+    {
+      "epoch": 0.1126229019763631,
+      "grad_norm": 0.5298368232940289,
+      "learning_rate": 9.822761620836158e-06,
+      "loss": 0.0177,
+      "step": 567
+    },
+    {
+      "epoch": 0.11282153143311154,
+      "grad_norm": 0.7938863386225405,
+      "learning_rate": 9.821911543960025e-06,
+      "loss": 0.0117,
+      "step": 568
+    },
+    {
+      "epoch": 0.11302016088985997,
+      "grad_norm": 0.8596537266570184,
+      "learning_rate": 9.82105947033515e-06,
+      "loss": 0.0072,
+      "step": 569
+    },
+    {
+      "epoch": 0.11321879034660841,
+      "grad_norm": 0.4173258192767521,
+      "learning_rate": 9.820205400314378e-06,
+      "loss": 0.0145,
+      "step": 570
+    },
+    {
+      "epoch": 0.11341741980335683,
+      "grad_norm": 0.6115201580608163,
+      "learning_rate": 9.819349334251376e-06,
+      "loss": 0.0203,
+      "step": 571
+    },
+    {
+      "epoch": 0.11361604926010527,
+      "grad_norm": 0.30050628933401596,
+      "learning_rate": 9.818491272500641e-06,
+      "loss": 0.0107,
+      "step": 572
+    },
+    {
+      "epoch": 0.1138146787168537,
+      "grad_norm": 1.0979707173146247,
+      "learning_rate": 9.817631215417496e-06,
+      "loss": 0.0223,
+      "step": 573
+    },
+    {
+      "epoch": 0.11401330817360214,
+      "grad_norm": 0.46276625355749734,
+      "learning_rate": 9.816769163358087e-06,
+      "loss": 0.0125,
+      "step": 574
+    },
+    {
+      "epoch": 0.11421193763035058,
+      "grad_norm": 1.0534382953635288,
+      "learning_rate": 9.815905116679394e-06,
+      "loss": 0.0239,
+      "step": 575
+    },
+    {
+      "epoch": 0.11441056708709901,
+      "grad_norm": 0.4199398914552984,
+      "learning_rate": 9.815039075739213e-06,
+      "loss": 0.0113,
+      "step": 576
+    },
+    {
+      "epoch": 0.11460919654384745,
+      "grad_norm": 1.1636761751492717,
+      "learning_rate": 9.814171040896173e-06,
+      "loss": 0.0208,
+      "step": 577
+    },
+    {
+      "epoch": 0.11480782600059589,
+      "grad_norm": 0.5065188833689531,
+      "learning_rate": 9.813301012509725e-06,
+      "loss": 0.0081,
+      "step": 578
+    },
+    {
+      "epoch": 0.11500645545734432,
+      "grad_norm": 1.1440102400398982,
+      "learning_rate": 9.812428990940151e-06,
+      "loss": 0.0201,
+      "step": 579
+    },
+    {
+      "epoch": 0.11520508491409276,
+      "grad_norm": 0.5140689221203485,
+      "learning_rate": 9.811554976548547e-06,
+      "loss": 0.0154,
+      "step": 580
+    },
+    {
+      "epoch": 0.1154037143708412,
+      "grad_norm": 0.8604730664877108,
+      "learning_rate": 9.810678969696849e-06,
+      "loss": 0.0216,
+      "step": 581
+    },
+    {
+      "epoch": 0.11560234382758963,
+      "grad_norm": 1.3627495098285676,
+      "learning_rate": 9.809800970747805e-06,
+      "loss": 0.0146,
+      "step": 582
+    },
+    {
+      "epoch": 0.11580097328433807,
+      "grad_norm": 1.5215024670202477,
+      "learning_rate": 9.808920980064998e-06,
+      "loss": 0.0187,
+      "step": 583
+    },
+    {
+      "epoch": 0.1159996027410865,
+      "grad_norm": 0.6734406716826478,
+      "learning_rate": 9.808038998012828e-06,
+      "loss": 0.0174,
+      "step": 584
+    },
+    {
+      "epoch": 0.11619823219783494,
+      "grad_norm": 1.1041085950914342,
+      "learning_rate": 9.807155024956523e-06,
+      "loss": 0.0151,
+      "step": 585
+    },
+    {
+      "epoch": 0.11639686165458338,
+      "grad_norm": 1.5387484405002336,
+      "learning_rate": 9.806269061262135e-06,
+      "loss": 0.0149,
+      "step": 586
+    },
+    {
+      "epoch": 0.11659549111133181,
+      "grad_norm": 0.5875043750348269,
+      "learning_rate": 9.805381107296544e-06,
+      "loss": 0.0143,
+      "step": 587
+    },
+    {
+      "epoch": 0.11679412056808025,
+      "grad_norm": 0.8389398799982417,
+      "learning_rate": 9.80449116342745e-06,
+      "loss": 0.0255,
+      "step": 588
+    },
+    {
+      "epoch": 0.11699275002482869,
+      "grad_norm": 0.30889573926063935,
+      "learning_rate": 9.803599230023373e-06,
+      "loss": 0.0103,
+      "step": 589
+    },
+    {
+      "epoch": 0.11719137948157712,
+      "grad_norm": 0.7219078584006366,
+      "learning_rate": 9.802705307453667e-06,
+      "loss": 0.0125,
+      "step": 590
+    },
+    {
+      "epoch": 0.11739000893832556,
+      "grad_norm": 0.5466172579541949,
+      "learning_rate": 9.801809396088501e-06,
+      "loss": 0.0116,
+      "step": 591
+    },
+    {
+      "epoch": 0.11758863839507398,
+      "grad_norm": 1.0372207668042581,
+      "learning_rate": 9.800911496298875e-06,
+      "loss": 0.0123,
+      "step": 592
+    },
+    {
+      "epoch": 0.11778726785182242,
+      "grad_norm": 0.7559866086265034,
+      "learning_rate": 9.800011608456607e-06,
+      "loss": 0.0201,
+      "step": 593
+    },
+    {
+      "epoch": 0.11798589730857086,
+      "grad_norm": 0.3188709050740313,
+      "learning_rate": 9.799109732934338e-06,
+      "loss": 0.0111,
+      "step": 594
+    },
+    {
+      "epoch": 0.11818452676531929,
+      "grad_norm": 0.5314042057902478,
+      "learning_rate": 9.798205870105533e-06,
+      "loss": 0.0125,
+      "step": 595
+    },
+    {
+      "epoch": 0.11838315622206773,
+      "grad_norm": 1.8474069548887428,
+      "learning_rate": 9.797300020344483e-06,
+      "loss": 0.0231,
+      "step": 596
+    },
+    {
+      "epoch": 0.11858178567881617,
+      "grad_norm": 0.42452537891107434,
+      "learning_rate": 9.796392184026298e-06,
+      "loss": 0.0108,
+      "step": 597
+    },
+    {
+      "epoch": 0.1187804151355646,
+      "grad_norm": 0.5783590129051792,
+      "learning_rate": 9.795482361526915e-06,
+      "loss": 0.0116,
+      "step": 598
+    },
+    {
+      "epoch": 0.11897904459231304,
+      "grad_norm": 0.38163168855460666,
+      "learning_rate": 9.794570553223087e-06,
+      "loss": 0.009,
+      "step": 599
+    },
+    {
+      "epoch": 0.11917767404906147,
+      "grad_norm": 1.4917576733893148,
+      "learning_rate": 9.793656759492394e-06,
+      "loss": 0.0155,
+      "step": 600
+    },
+    {
+      "epoch": 0.11937630350580991,
+      "grad_norm": 1.0584955957981348,
+      "learning_rate": 9.792740980713238e-06,
+      "loss": 0.0207,
+      "step": 601
+    },
+    {
+      "epoch": 0.11957493296255835,
+      "grad_norm": 0.694913429075098,
+      "learning_rate": 9.791823217264842e-06,
+      "loss": 0.0161,
+      "step": 602
+    },
+    {
+      "epoch": 0.11977356241930678,
+      "grad_norm": 0.6695725185833449,
+      "learning_rate": 9.790903469527249e-06,
+      "loss": 0.0199,
+      "step": 603
+    },
+    {
+      "epoch": 0.11997219187605522,
+      "grad_norm": 2.000789888070613,
+      "learning_rate": 9.789981737881326e-06,
+      "loss": 0.039,
+      "step": 604
+    },
+    {
+      "epoch": 0.12017082133280366,
+      "grad_norm": 0.48955671291310315,
+      "learning_rate": 9.789058022708765e-06,
+      "loss": 0.0142,
+      "step": 605
+    },
+    {
+      "epoch": 0.1203694507895521,
+      "grad_norm": 1.0363261312441616,
+      "learning_rate": 9.788132324392072e-06,
+      "loss": 0.0159,
+      "step": 606
+    },
+    {
+      "epoch": 0.12056808024630053,
+      "grad_norm": 0.4884275959706906,
+      "learning_rate": 9.787204643314577e-06,
+      "loss": 0.0167,
+      "step": 607
+    },
+    {
+      "epoch": 0.12076670970304897,
+      "grad_norm": 0.7350563044633582,
+      "learning_rate": 9.786274979860434e-06,
+      "loss": 0.0158,
+      "step": 608
+    },
+    {
+      "epoch": 0.1209653391597974,
+      "grad_norm": 0.9892060951239566,
+      "learning_rate": 9.785343334414615e-06,
+      "loss": 0.0206,
+      "step": 609
+    },
+    {
+      "epoch": 0.12116396861654584,
+      "grad_norm": 0.37127352309370415,
+      "learning_rate": 9.784409707362913e-06,
+      "loss": 0.0183,
+      "step": 610
+    },
+    {
+      "epoch": 0.12136259807329428,
+      "grad_norm": 0.650383126477588,
+      "learning_rate": 9.783474099091943e-06,
+      "loss": 0.0176,
+      "step": 611
+    },
+    {
+      "epoch": 0.12156122753004271,
+      "grad_norm": 0.4119293084929478,
+      "learning_rate": 9.78253650998914e-06,
+      "loss": 0.0117,
+      "step": 612
+    },
+    {
+      "epoch": 0.12175985698679113,
+      "grad_norm": 0.829964935039823,
+      "learning_rate": 9.781596940442755e-06,
+      "loss": 0.0153,
+      "step": 613
+    },
+    {
+      "epoch": 0.12195848644353957,
+      "grad_norm": 1.7129449944097686,
+      "learning_rate": 9.780655390841867e-06,
+      "loss": 0.0136,
+      "step": 614
+    },
+    {
+      "epoch": 0.12215711590028801,
+      "grad_norm": 0.5565425099390852,
+      "learning_rate": 9.779711861576368e-06,
+      "loss": 0.0215,
+      "step": 615
+    },
+    {
+      "epoch": 0.12235574535703644,
+      "grad_norm": 0.7788538582287662,
+      "learning_rate": 9.778766353036975e-06,
+      "loss": 0.0218,
+      "step": 616
+    },
+    {
+      "epoch": 0.12255437481378488,
+      "grad_norm": 0.33444210785480794,
+      "learning_rate": 9.777818865615221e-06,
+      "loss": 0.0136,
+      "step": 617
+    },
+    {
+      "epoch": 0.12275300427053332,
+      "grad_norm": 0.400994084696202,
+      "learning_rate": 9.776869399703458e-06,
+      "loss": 0.0154,
+      "step": 618
+    },
+    {
+      "epoch": 0.12295163372728175,
+      "grad_norm": 0.7796516460912207,
+      "learning_rate": 9.77591795569486e-06,
+      "loss": 0.013,
+      "step": 619
+    },
+    {
+      "epoch": 0.12315026318403019,
+      "grad_norm": 0.536772489299672,
+      "learning_rate": 9.774964533983421e-06,
+      "loss": 0.0188,
+      "step": 620
+    },
+    {
+      "epoch": 0.12334889264077863,
+      "grad_norm": 1.3303202817826634,
+      "learning_rate": 9.774009134963949e-06,
+      "loss": 0.0193,
+      "step": 621
+    },
+    {
+      "epoch": 0.12354752209752706,
+      "grad_norm": 0.7037843214385814,
+      "learning_rate": 9.773051759032074e-06,
+      "loss": 0.0175,
+      "step": 622
+    },
+    {
+      "epoch": 0.1237461515542755,
+      "grad_norm": 0.4308203918294435,
+      "learning_rate": 9.772092406584248e-06,
+      "loss": 0.0166,
+      "step": 623
+    },
+    {
+      "epoch": 0.12394478101102394,
+      "grad_norm": 1.1697494943206082,
+      "learning_rate": 9.771131078017735e-06,
+      "loss": 0.0193,
+      "step": 624
+    },
+    {
+      "epoch": 0.12414341046777237,
+      "grad_norm": 0.5812375266471298,
+      "learning_rate": 9.77016777373062e-06,
+      "loss": 0.0146,
+      "step": 625
+    },
+    {
+      "epoch": 0.12434203992452081,
+      "grad_norm": 0.524198461309644,
+      "learning_rate": 9.769202494121806e-06,
+      "loss": 0.0167,
+      "step": 626
+    },
+    {
+      "epoch": 0.12454066938126925,
+      "grad_norm": 1.5285962472865722,
+      "learning_rate": 9.768235239591016e-06,
+      "loss": 0.0207,
+      "step": 627
+    },
+    {
+      "epoch": 0.12473929883801768,
+      "grad_norm": 0.7396830974883195,
+      "learning_rate": 9.767266010538788e-06,
+      "loss": 0.0148,
+      "step": 628
+    },
+    {
+      "epoch": 0.12493792829476612,
+      "grad_norm": 0.2274647843251192,
+      "learning_rate": 9.76629480736648e-06,
+      "loss": 0.0107,
+      "step": 629
+    },
+    {
+      "epoch": 0.12513655775151455,
+      "grad_norm": 0.28859531186159604,
+      "learning_rate": 9.765321630476264e-06,
+      "loss": 0.0166,
+      "step": 630
+    },
+    {
+      "epoch": 0.12533518720826298,
+      "grad_norm": 0.35440823267343124,
+      "learning_rate": 9.764346480271132e-06,
+      "loss": 0.0168,
+      "step": 631
+    },
+    {
+      "epoch": 0.12553381666501143,
+      "grad_norm": 0.5712874545330476,
+      "learning_rate": 9.763369357154895e-06,
+      "loss": 0.0101,
+      "step": 632
+    },
+    {
+      "epoch": 0.12573244612175985,
+      "grad_norm": 1.3923546120285113,
+      "learning_rate": 9.762390261532177e-06,
+      "loss": 0.0168,
+      "step": 633
+    },
+    {
+      "epoch": 0.1259310755785083,
+      "grad_norm": 0.47850667989496215,
+      "learning_rate": 9.761409193808422e-06,
+      "loss": 0.009,
+      "step": 634
+    },
+    {
+      "epoch": 0.12612970503525672,
+      "grad_norm": 0.4045494692733913,
+      "learning_rate": 9.760426154389888e-06,
+      "loss": 0.0127,
+      "step": 635
+    },
+    {
+      "epoch": 0.12632833449200517,
+      "grad_norm": 0.949458135135656,
+      "learning_rate": 9.75944114368365e-06,
+      "loss": 0.0245,
+      "step": 636
+    },
+    {
+      "epoch": 0.1265269639487536,
+      "grad_norm": 0.4744219220416367,
+      "learning_rate": 9.7584541620976e-06,
+      "loss": 0.01,
+      "step": 637
+    },
+    {
+      "epoch": 0.12672559340550205,
+      "grad_norm": 0.6580474069905136,
+      "learning_rate": 9.757465210040447e-06,
+      "loss": 0.0116,
+      "step": 638
+    },
+    {
+      "epoch": 0.12692422286225047,
+      "grad_norm": 1.2746603803218375,
+      "learning_rate": 9.756474287921716e-06,
+      "loss": 0.0173,
+      "step": 639
+    },
+    {
+      "epoch": 0.12712285231899892,
+      "grad_norm": 1.9417494348440496,
+      "learning_rate": 9.755481396151744e-06,
+      "loss": 0.0324,
+      "step": 640
+    },
+    {
+      "epoch": 0.12732148177574734,
+      "grad_norm": 0.618531205863786,
+      "learning_rate": 9.75448653514169e-06,
+      "loss": 0.0142,
+      "step": 641
+    },
+    {
+      "epoch": 0.1275201112324958,
+      "grad_norm": 0.9971146664077777,
+      "learning_rate": 9.75348970530352e-06,
+      "loss": 0.018,
+      "step": 642
+    },
+    {
+      "epoch": 0.12771874068924421,
+      "grad_norm": 0.8389497013907814,
+      "learning_rate": 9.752490907050027e-06,
+      "loss": 0.0151,
+      "step": 643
+    },
+    {
+      "epoch": 0.12791737014599264,
+      "grad_norm": 1.0391641435545595,
+      "learning_rate": 9.751490140794807e-06,
+      "loss": 0.0234,
+      "step": 644
+    },
+    {
+      "epoch": 0.1281159996027411,
+      "grad_norm": 0.7118837598081941,
+      "learning_rate": 9.750487406952277e-06,
+      "loss": 0.0165,
+      "step": 645
+    },
+    {
+      "epoch": 0.1283146290594895,
+      "grad_norm": 0.3126685428380962,
+      "learning_rate": 9.74948270593767e-06,
+      "loss": 0.0104,
+      "step": 646
+    },
+    {
+      "epoch": 0.12851325851623796,
+      "grad_norm": 0.6400960063705387,
+      "learning_rate": 9.74847603816703e-06,
+      "loss": 0.0146,
+      "step": 647
+    },
+    {
+      "epoch": 0.12871188797298638,
+      "grad_norm": 0.6080027304759789,
+      "learning_rate": 9.747467404057219e-06,
+      "loss": 0.0132,
+      "step": 648
+    },
+    {
+      "epoch": 0.12891051742973483,
+      "grad_norm": 0.5969440680204644,
+      "learning_rate": 9.74645680402591e-06,
+      "loss": 0.0165,
+      "step": 649
+    },
+    {
+      "epoch": 0.12910914688648326,
+      "grad_norm": 0.4844450633614373,
+      "learning_rate": 9.74544423849159e-06,
+      "loss": 0.008,
+      "step": 650
+    },
+    {
+      "epoch": 0.1293077763432317,
+      "grad_norm": 0.5731462801181816,
+      "learning_rate": 9.744429707873564e-06,
+      "loss": 0.014,
+      "step": 651
+    },
+    {
+      "epoch": 0.12950640579998013,
+      "grad_norm": 0.9087700604835134,
+      "learning_rate": 9.743413212591949e-06,
+      "loss": 0.0213,
+      "step": 652
+    },
+    {
+      "epoch": 0.12970503525672858,
+      "grad_norm": 0.7602480663063869,
+      "learning_rate": 9.742394753067671e-06,
+      "loss": 0.0158,
+      "step": 653
+    },
+    {
+      "epoch": 0.129903664713477,
+      "grad_norm": 0.6559403776033135,
+      "learning_rate": 9.741374329722474e-06,
+      "loss": 0.0126,
+      "step": 654
+    },
+    {
+      "epoch": 0.13010229417022545,
+      "grad_norm": 0.5644269287564567,
+      "learning_rate": 9.740351942978919e-06,
+      "loss": 0.0141,
+      "step": 655
+    },
+    {
+      "epoch": 0.13030092362697387,
+      "grad_norm": 0.5606392709106318,
+      "learning_rate": 9.739327593260367e-06,
+      "loss": 0.0221,
+      "step": 656
+    },
+    {
+      "epoch": 0.13049955308372232,
+      "grad_norm": 0.6807347976418896,
+      "learning_rate": 9.738301280991007e-06,
+      "loss": 0.0151,
+      "step": 657
+    },
+    {
+      "epoch": 0.13069818254047075,
+      "grad_norm": 0.8462663632413303,
+      "learning_rate": 9.737273006595832e-06,
+      "loss": 0.0194,
+      "step": 658
+    },
+    {
+      "epoch": 0.1308968119972192,
+      "grad_norm": 0.3894055488420919,
+      "learning_rate": 9.736242770500647e-06,
+      "loss": 0.0092,
+      "step": 659
+    },
+    {
+      "epoch": 0.13109544145396762,
+      "grad_norm": 0.7904884629984615,
+      "learning_rate": 9.735210573132078e-06,
+      "loss": 0.0162,
+      "step": 660
+    },
+    {
+      "epoch": 0.13129407091071607,
+      "grad_norm": 0.9674854160105776,
+      "learning_rate": 9.734176414917548e-06,
+      "loss": 0.0196,
+      "step": 661
+    },
+    {
+      "epoch": 0.1314927003674645,
+      "grad_norm": 0.33200341815567036,
+      "learning_rate": 9.733140296285307e-06,
+      "loss": 0.0093,
+      "step": 662
+    },
+    {
+      "epoch": 0.13169132982421294,
+      "grad_norm": 0.8720779831059587,
+      "learning_rate": 9.73210221766441e-06,
+      "loss": 0.0171,
+      "step": 663
+    },
+    {
+      "epoch": 0.13188995928096137,
+      "grad_norm": 0.48679544383358947,
+      "learning_rate": 9.731062179484723e-06,
+      "loss": 0.0222,
+      "step": 664
+    },
+    {
+      "epoch": 0.1320885887377098,
+      "grad_norm": 1.6442374114676053,
+      "learning_rate": 9.730020182176925e-06,
+      "loss": 0.0201,
+      "step": 665
+    },
+    {
+      "epoch": 0.13228721819445824,
+      "grad_norm": 0.6211726152538897,
+      "learning_rate": 9.728976226172507e-06,
+      "loss": 0.0115,
+      "step": 666
+    },
+    {
+      "epoch": 0.13248584765120666,
+      "grad_norm": 1.0316344031063414,
+      "learning_rate": 9.727930311903768e-06,
+      "loss": 0.0138,
+      "step": 667
+    },
+    {
+      "epoch": 0.1326844771079551,
+      "grad_norm": 0.6585612814093009,
+      "learning_rate": 9.726882439803822e-06,
+      "loss": 0.0221,
+      "step": 668
+    },
+    {
+      "epoch": 0.13288310656470353,
+      "grad_norm": 1.583096886947926,
+      "learning_rate": 9.725832610306592e-06,
+      "loss": 0.0236,
+      "step": 669
+    },
+    {
+      "epoch": 0.13308173602145199,
+      "grad_norm": 0.58785647706293,
+      "learning_rate": 9.72478082384681e-06,
+      "loss": 0.0133,
+      "step": 670
+    },
+    {
+      "epoch": 0.1332803654782004,
+      "grad_norm": 0.4215741160538269,
+      "learning_rate": 9.723727080860022e-06,
+      "loss": 0.0117,
+      "step": 671
+    },
+    {
+      "epoch": 0.13347899493494886,
+      "grad_norm": 1.1714886610581292,
+      "learning_rate": 9.722671381782577e-06,
+      "loss": 0.0289,
+      "step": 672
+    },
+    {
+      "epoch": 0.13367762439169728,
+      "grad_norm": 0.849483842098239,
+      "learning_rate": 9.721613727051646e-06,
+      "loss": 0.017,
+      "step": 673
+    },
+    {
+      "epoch": 0.13387625384844573,
+      "grad_norm": 1.4736074773129628,
+      "learning_rate": 9.720554117105197e-06,
+      "loss": 0.0132,
+      "step": 674
+    },
+    {
+      "epoch": 0.13407488330519415,
+      "grad_norm": 2.2121758689867344,
+      "learning_rate": 9.719492552382015e-06,
+      "loss": 0.0268,
+      "step": 675
+    },
+    {
+      "epoch": 0.1342735127619426,
+      "grad_norm": 2.2633647738559546,
+      "learning_rate": 9.718429033321693e-06,
+      "loss": 0.0279,
+      "step": 676
+    },
+    {
+      "epoch": 0.13447214221869103,
+      "grad_norm": 0.5748397779881033,
+      "learning_rate": 9.717363560364634e-06,
+      "loss": 0.0198,
+      "step": 677
+    },
+    {
+      "epoch": 0.13467077167543948,
+      "grad_norm": 0.7319060757746011,
+      "learning_rate": 9.71629613395205e-06,
+      "loss": 0.015,
+      "step": 678
+    },
+    {
+      "epoch": 0.1348694011321879,
+      "grad_norm": 0.8521409511987582,
+      "learning_rate": 9.71522675452596e-06,
+      "loss": 0.0141,
+      "step": 679
+    },
+    {
+      "epoch": 0.13506803058893635,
+      "grad_norm": 0.910962721954014,
+      "learning_rate": 9.714155422529192e-06,
+      "loss": 0.016,
+      "step": 680
+    },
+    {
+      "epoch": 0.13526666004568477,
+      "grad_norm": 0.3260897607925365,
+      "learning_rate": 9.713082138405383e-06,
+      "loss": 0.0105,
+      "step": 681
+    },
+    {
+      "epoch": 0.13546528950243322,
+      "grad_norm": 0.2535946790266379,
+      "learning_rate": 9.712006902598982e-06,
+      "loss": 0.0112,
+      "step": 682
+    },
+    {
+      "epoch": 0.13566391895918165,
+      "grad_norm": 0.4071046693159764,
+      "learning_rate": 9.710929715555241e-06,
+      "loss": 0.0171,
+      "step": 683
+    },
+    {
+      "epoch": 0.1358625484159301,
+      "grad_norm": 0.6842960302160229,
+      "learning_rate": 9.709850577720223e-06,
+      "loss": 0.0123,
+      "step": 684
+    },
+    {
+      "epoch": 0.13606117787267852,
+      "grad_norm": 0.35348541207601764,
+      "learning_rate": 9.708769489540796e-06,
+      "loss": 0.0127,
+      "step": 685
+    },
+    {
+      "epoch": 0.13625980732942694,
+      "grad_norm": 0.5960992242072756,
+      "learning_rate": 9.70768645146464e-06,
+      "loss": 0.0152,
+      "step": 686
+    },
+    {
+      "epoch": 0.1364584367861754,
+      "grad_norm": 0.706647286718095,
+      "learning_rate": 9.706601463940237e-06,
+      "loss": 0.0092,
+      "step": 687
+    },
+    {
+      "epoch": 0.1366570662429238,
+      "grad_norm": 0.8393587614699919,
+      "learning_rate": 9.705514527416885e-06,
+      "loss": 0.0167,
+      "step": 688
+    },
+    {
+      "epoch": 0.13685569569967226,
+      "grad_norm": 0.8895837213581118,
+      "learning_rate": 9.704425642344674e-06,
+      "loss": 0.017,
+      "step": 689
+    },
+    {
+      "epoch": 0.1370543251564207,
+      "grad_norm": 0.6798384157696477,
+      "learning_rate": 9.703334809174519e-06,
+      "loss": 0.0117,
+      "step": 690
+    },
+    {
+      "epoch": 0.13725295461316914,
+      "grad_norm": 0.7053098557106127,
+      "learning_rate": 9.70224202835813e-06,
+      "loss": 0.0131,
+      "step": 691
+    },
+    {
+      "epoch": 0.13745158406991756,
+      "grad_norm": 1.1725400512260025,
+      "learning_rate": 9.701147300348025e-06,
+      "loss": 0.0133,
+      "step": 692
+    },
+    {
+      "epoch": 0.137650213526666,
+      "grad_norm": 0.4166073254333175,
+      "learning_rate": 9.70005062559753e-06,
+      "loss": 0.0124,
+      "step": 693
+    },
+    {
+      "epoch": 0.13784884298341443,
+      "grad_norm": 0.5431686497693445,
+      "learning_rate": 9.69895200456078e-06,
+      "loss": 0.0079,
+      "step": 694
+    },
+    {
+      "epoch": 0.13804747244016288,
+      "grad_norm": 0.5724175575189289,
+      "learning_rate": 9.697851437692708e-06,
+      "loss": 0.0172,
+      "step": 695
+    },
+    {
+      "epoch": 0.1382461018969113,
+      "grad_norm": 1.243225503850031,
+      "learning_rate": 9.696748925449061e-06,
+      "loss": 0.0225,
+      "step": 696
+    },
+    {
+      "epoch": 0.13844473135365976,
+      "grad_norm": 0.5911282831149057,
+      "learning_rate": 9.69564446828639e-06,
+      "loss": 0.0167,
+      "step": 697
+    },
+    {
+      "epoch": 0.13864336081040818,
+      "grad_norm": 0.7689274476457225,
+      "learning_rate": 9.694538066662043e-06,
+      "loss": 0.0165,
+      "step": 698
+    },
+    {
+      "epoch": 0.13884199026715663,
+      "grad_norm": 0.9634756763078826,
+      "learning_rate": 9.693429721034186e-06,
+      "loss": 0.0205,
+      "step": 699
+    },
+    {
+      "epoch": 0.13904061972390505,
+      "grad_norm": 0.9777879206227595,
+      "learning_rate": 9.69231943186178e-06,
+      "loss": 0.0138,
+      "step": 700
+    },
+    {
+      "epoch": 0.1392392491806535,
+      "grad_norm": 0.5707455498722125,
+      "learning_rate": 9.691207199604599e-06,
+      "loss": 0.0191,
+      "step": 701
+    },
+    {
+      "epoch": 0.13943787863740192,
+      "grad_norm": 1.3726637583023185,
+      "learning_rate": 9.690093024723213e-06,
+      "loss": 0.0165,
+      "step": 702
+    },
+    {
+      "epoch": 0.13963650809415037,
+      "grad_norm": 1.1668665383346433,
+      "learning_rate": 9.688976907679001e-06,
+      "loss": 0.0174,
+      "step": 703
+    },
+    {
+      "epoch": 0.1398351375508988,
+      "grad_norm": 1.0967041976393477,
+      "learning_rate": 9.68785884893415e-06,
+      "loss": 0.0161,
+      "step": 704
+    },
+    {
+      "epoch": 0.14003376700764725,
+      "grad_norm": 0.3589397157463669,
+      "learning_rate": 9.686738848951642e-06,
+      "loss": 0.0226,
+      "step": 705
+    },
+    {
+      "epoch": 0.14023239646439567,
+      "grad_norm": 0.37386431956811017,
+      "learning_rate": 9.68561690819527e-06,
+      "loss": 0.012,
+      "step": 706
+    },
+    {
+      "epoch": 0.1404310259211441,
+      "grad_norm": 1.3981607309328512,
+      "learning_rate": 9.68449302712963e-06,
+      "loss": 0.0167,
+      "step": 707
+    },
+    {
+      "epoch": 0.14062965537789254,
+      "grad_norm": 1.1763685062897045,
+      "learning_rate": 9.683367206220118e-06,
+      "loss": 0.0224,
+      "step": 708
+    },
+    {
+      "epoch": 0.14082828483464097,
+      "grad_norm": 1.943944629520424,
+      "learning_rate": 9.682239445932937e-06,
+      "loss": 0.0272,
+      "step": 709
+    },
+    {
+      "epoch": 0.14102691429138942,
+      "grad_norm": 0.6898744285476524,
+      "learning_rate": 9.681109746735089e-06,
+      "loss": 0.0142,
+      "step": 710
+    },
+    {
+      "epoch": 0.14122554374813784,
+      "grad_norm": 0.5450517944671093,
+      "learning_rate": 9.679978109094383e-06,
+      "loss": 0.0169,
+      "step": 711
+    },
+    {
+      "epoch": 0.1414241732048863,
+      "grad_norm": 0.44279377105851336,
+      "learning_rate": 9.678844533479427e-06,
+      "loss": 0.013,
+      "step": 712
+    },
+    {
+      "epoch": 0.1416228026616347,
+      "grad_norm": 1.6342810458603632,
+      "learning_rate": 9.677709020359638e-06,
+      "loss": 0.0254,
+      "step": 713
+    },
+    {
+      "epoch": 0.14182143211838316,
+      "grad_norm": 2.0291941877579065,
+      "learning_rate": 9.676571570205227e-06,
+      "loss": 0.0119,
+      "step": 714
+    },
+    {
+      "epoch": 0.14202006157513158,
+      "grad_norm": 1.5157380559444824,
+      "learning_rate": 9.675432183487211e-06,
+      "loss": 0.0162,
+      "step": 715
+    },
+    {
+      "epoch": 0.14221869103188003,
+      "grad_norm": 1.0513407088044642,
+      "learning_rate": 9.67429086067741e-06,
+      "loss": 0.0202,
+      "step": 716
+    },
+    {
+      "epoch": 0.14241732048862846,
+      "grad_norm": 0.8011495579478646,
+      "learning_rate": 9.673147602248448e-06,
+      "loss": 0.0165,
+      "step": 717
+    },
+    {
+      "epoch": 0.1426159499453769,
+      "grad_norm": 0.41050875738974985,
+      "learning_rate": 9.67200240867374e-06,
+      "loss": 0.0145,
+      "step": 718
+    },
+    {
+      "epoch": 0.14281457940212533,
+      "grad_norm": 1.2640264053922001,
+      "learning_rate": 9.670855280427514e-06,
+      "loss": 0.0157,
+      "step": 719
+    },
+    {
+      "epoch": 0.14301320885887378,
+      "grad_norm": 0.7857274636818972,
+      "learning_rate": 9.669706217984793e-06,
+      "loss": 0.0105,
+      "step": 720
+    },
+    {
+      "epoch": 0.1432118383156222,
+      "grad_norm": 0.6636140885305981,
+      "learning_rate": 9.668555221821404e-06,
+      "loss": 0.0132,
+      "step": 721
+    },
+    {
+      "epoch": 0.14341046777237065,
+      "grad_norm": 0.6949080113239285,
+      "learning_rate": 9.667402292413975e-06,
+      "loss": 0.0131,
+      "step": 722
+    },
+    {
+      "epoch": 0.14360909722911908,
+      "grad_norm": 0.7348391840249473,
+      "learning_rate": 9.66624743023993e-06,
+      "loss": 0.0135,
+      "step": 723
+    },
+    {
+      "epoch": 0.14380772668586753,
+      "grad_norm": 0.515623939890902,
+      "learning_rate": 9.665090635777497e-06,
+      "loss": 0.0171,
+      "step": 724
+    },
+    {
+      "epoch": 0.14400635614261595,
+      "grad_norm": 0.8344063986031619,
+      "learning_rate": 9.663931909505702e-06,
+      "loss": 0.023,
+      "step": 725
+    },
+    {
+      "epoch": 0.1442049855993644,
+      "grad_norm": 1.6382892707257288,
+      "learning_rate": 9.662771251904375e-06,
+      "loss": 0.0214,
+      "step": 726
+    },
+    {
+      "epoch": 0.14440361505611282,
+      "grad_norm": 0.6588029713890065,
+      "learning_rate": 9.66160866345414e-06,
+      "loss": 0.0092,
+      "step": 727
+    },
+    {
+      "epoch": 0.14460224451286124,
+      "grad_norm": 0.5348463144307005,
+      "learning_rate": 9.660444144636429e-06,
+      "loss": 0.0084,
+      "step": 728
+    },
+    {
+      "epoch": 0.1448008739696097,
+      "grad_norm": 0.765885295893399,
+      "learning_rate": 9.659277695933462e-06,
+      "loss": 0.0175,
+      "step": 729
+    },
+    {
+      "epoch": 0.14499950342635812,
+      "grad_norm": 0.8497299979434756,
+      "learning_rate": 9.658109317828267e-06,
+      "loss": 0.019,
+      "step": 730
+    },
+    {
+      "epoch": 0.14519813288310657,
+      "grad_norm": 0.523328490867962,
+      "learning_rate": 9.656939010804672e-06,
+      "loss": 0.0216,
+      "step": 731
+    },
+    {
+      "epoch": 0.145396762339855,
+      "grad_norm": 0.6009169743511245,
+      "learning_rate": 9.655766775347292e-06,
+      "loss": 0.0125,
+      "step": 732
+    },
+    {
+      "epoch": 0.14559539179660344,
+      "grad_norm": 0.6220910948352019,
+      "learning_rate": 9.654592611941555e-06,
+      "loss": 0.014,
+      "step": 733
+    },
+    {
+      "epoch": 0.14579402125335186,
+      "grad_norm": 0.7443834281999502,
+      "learning_rate": 9.653416521073678e-06,
+      "loss": 0.0118,
+      "step": 734
+    },
+    {
+      "epoch": 0.1459926507101003,
+      "grad_norm": 0.3777328532342268,
+      "learning_rate": 9.65223850323068e-06,
+      "loss": 0.016,
+      "step": 735
+    },
+    {
+      "epoch": 0.14619128016684874,
+      "grad_norm": 0.36374649077029947,
+      "learning_rate": 9.651058558900375e-06,
+      "loss": 0.0134,
+      "step": 736
+    },
+    {
+      "epoch": 0.1463899096235972,
+      "grad_norm": 0.343215847483735,
+      "learning_rate": 9.64987668857138e-06,
+      "loss": 0.0175,
+      "step": 737
+    },
+    {
+      "epoch": 0.1465885390803456,
+      "grad_norm": 0.38715395916519696,
+      "learning_rate": 9.648692892733105e-06,
+      "loss": 0.0085,
+      "step": 738
+    },
+    {
+      "epoch": 0.14678716853709406,
+      "grad_norm": 0.27505352925401316,
+      "learning_rate": 9.647507171875758e-06,
+      "loss": 0.0064,
+      "step": 739
+    },
+    {
+      "epoch": 0.14698579799384248,
+      "grad_norm": 0.7665979379551244,
+      "learning_rate": 9.646319526490345e-06,
+      "loss": 0.0185,
+      "step": 740
+    },
+    {
+      "epoch": 0.14718442745059093,
+      "grad_norm": 0.4034771698141865,
+      "learning_rate": 9.64512995706867e-06,
+      "loss": 0.0114,
+      "step": 741
+    },
+    {
+      "epoch": 0.14738305690733935,
+      "grad_norm": 0.6233753352700463,
+      "learning_rate": 9.643938464103331e-06,
+      "loss": 0.0205,
+      "step": 742
+    },
+    {
+      "epoch": 0.1475816863640878,
+      "grad_norm": 0.6387978330419878,
+      "learning_rate": 9.642745048087724e-06,
+      "loss": 0.0122,
+      "step": 743
+    },
+    {
+      "epoch": 0.14778031582083623,
+      "grad_norm": 0.6530117233777584,
+      "learning_rate": 9.641549709516042e-06,
+      "loss": 0.0242,
+      "step": 744
+    },
+    {
+      "epoch": 0.14797894527758468,
+      "grad_norm": 0.6726282197782968,
+      "learning_rate": 9.640352448883273e-06,
+      "loss": 0.0207,
+      "step": 745
+    },
+    {
+      "epoch": 0.1481775747343331,
+      "grad_norm": 0.3645759942166644,
+      "learning_rate": 9.639153266685204e-06,
+      "loss": 0.0145,
+      "step": 746
+    },
+    {
+      "epoch": 0.14837620419108155,
+      "grad_norm": 0.4159658079352318,
+      "learning_rate": 9.63795216341841e-06,
+      "loss": 0.0167,
+      "step": 747
+    },
+    {
+      "epoch": 0.14857483364782997,
+      "grad_norm": 0.4271126375806547,
+      "learning_rate": 9.636749139580272e-06,
+      "loss": 0.0138,
+      "step": 748
+    },
+    {
+      "epoch": 0.1487734631045784,
+      "grad_norm": 0.659027659352402,
+      "learning_rate": 9.635544195668958e-06,
+      "loss": 0.0156,
+      "step": 749
+    },
+    {
+      "epoch": 0.14897209256132685,
+      "grad_norm": 0.42458273885035736,
+      "learning_rate": 9.634337332183435e-06,
+      "loss": 0.0193,
+      "step": 750
+    },
+    {
+      "epoch": 0.14917072201807527,
+      "grad_norm": 0.8691539370190741,
+      "learning_rate": 9.633128549623463e-06,
+      "loss": 0.0159,
+      "step": 751
+    },
+    {
+      "epoch": 0.14936935147482372,
+      "grad_norm": 0.8763737396333379,
+      "learning_rate": 9.6319178484896e-06,
+      "loss": 0.0174,
+      "step": 752
+    },
+    {
+      "epoch": 0.14956798093157214,
+      "grad_norm": 0.8359131444320231,
+      "learning_rate": 9.630705229283192e-06,
+      "loss": 0.0162,
+      "step": 753
+    },
+    {
+      "epoch": 0.1497666103883206,
+      "grad_norm": 0.3581978948784169,
+      "learning_rate": 9.629490692506386e-06,
+      "loss": 0.0171,
+      "step": 754
+    },
+    {
+      "epoch": 0.14996523984506901,
+      "grad_norm": 0.7735573422661927,
+      "learning_rate": 9.628274238662124e-06,
+      "loss": 0.0177,
+      "step": 755
+    },
+    {
+      "epoch": 0.15016386930181747,
+      "grad_norm": 0.6569012717678822,
+      "learning_rate": 9.627055868254131e-06,
+      "loss": 0.0165,
+      "step": 756
+    },
+    {
+      "epoch": 0.1503624987585659,
+      "grad_norm": 0.8802984939901024,
+      "learning_rate": 9.625835581786937e-06,
+      "loss": 0.0167,
+      "step": 757
+    },
+    {
+      "epoch": 0.15056112821531434,
+      "grad_norm": 0.42465561361686077,
+      "learning_rate": 9.624613379765863e-06,
+      "loss": 0.0112,
+      "step": 758
+    },
+    {
+      "epoch": 0.15075975767206276,
+      "grad_norm": 0.5587945392733407,
+      "learning_rate": 9.623389262697018e-06,
+      "loss": 0.0146,
+      "step": 759
+    },
+    {
+      "epoch": 0.1509583871288112,
+      "grad_norm": 0.6613542353383493,
+      "learning_rate": 9.62216323108731e-06,
+      "loss": 0.0091,
+      "step": 760
+    },
+    {
+      "epoch": 0.15115701658555963,
+      "grad_norm": 0.6745543946977648,
+      "learning_rate": 9.620935285444435e-06,
+      "loss": 0.0102,
+      "step": 761
+    },
+    {
+      "epoch": 0.15135564604230808,
+      "grad_norm": 1.0899874100297828,
+      "learning_rate": 9.619705426276887e-06,
+      "loss": 0.0101,
+      "step": 762
+    },
+    {
+      "epoch": 0.1515542754990565,
+      "grad_norm": 0.9734431138507607,
+      "learning_rate": 9.61847365409395e-06,
+      "loss": 0.0201,
+      "step": 763
+    },
+    {
+      "epoch": 0.15175290495580496,
+      "grad_norm": 0.6620377606505773,
+      "learning_rate": 9.617239969405696e-06,
+      "loss": 0.0108,
+      "step": 764
+    },
+    {
+      "epoch": 0.15195153441255338,
+      "grad_norm": 0.6543556889174846,
+      "learning_rate": 9.616004372722993e-06,
+      "loss": 0.0089,
+      "step": 765
+    },
+    {
+      "epoch": 0.15215016386930183,
+      "grad_norm": 1.1176456576818978,
+      "learning_rate": 9.614766864557505e-06,
+      "loss": 0.0149,
+      "step": 766
+    },
+    {
+      "epoch": 0.15234879332605025,
+      "grad_norm": 1.4536512494283313,
+      "learning_rate": 9.613527445421678e-06,
+      "loss": 0.0236,
+      "step": 767
+    },
+    {
+      "epoch": 0.1525474227827987,
+      "grad_norm": 0.9953239414430083,
+      "learning_rate": 9.612286115828757e-06,
+      "loss": 0.017,
+      "step": 768
+    },
+    {
+      "epoch": 0.15274605223954713,
+      "grad_norm": 0.9938930956416363,
+      "learning_rate": 9.611042876292774e-06,
+      "loss": 0.0167,
+      "step": 769
+    },
+    {
+      "epoch": 0.15294468169629555,
+      "grad_norm": 1.1047738551365187,
+      "learning_rate": 9.609797727328553e-06,
+      "loss": 0.0218,
+      "step": 770
+    },
+    {
+      "epoch": 0.153143311153044,
+      "grad_norm": 0.6023972990227837,
+      "learning_rate": 9.608550669451709e-06,
+      "loss": 0.006,
+      "step": 771
+    },
+    {
+      "epoch": 0.15334194060979242,
+      "grad_norm": 0.5085628473524939,
+      "learning_rate": 9.607301703178648e-06,
+      "loss": 0.0237,
+      "step": 772
+    },
+    {
+      "epoch": 0.15354057006654087,
+      "grad_norm": 0.7921029494776086,
+      "learning_rate": 9.606050829026568e-06,
+      "loss": 0.013,
+      "step": 773
+    },
+    {
+      "epoch": 0.1537391995232893,
+      "grad_norm": 0.3846454240514661,
+      "learning_rate": 9.604798047513449e-06,
+      "loss": 0.01,
+      "step": 774
+    },
+    {
+      "epoch": 0.15393782898003774,
+      "grad_norm": 0.8493324370898936,
+      "learning_rate": 9.603543359158071e-06,
+      "loss": 0.0206,
+      "step": 775
+    },
+    {
+      "epoch": 0.15413645843678617,
+      "grad_norm": 1.3234666263493877,
+      "learning_rate": 9.60228676448e-06,
+      "loss": 0.0173,
+      "step": 776
+    },
+    {
+      "epoch": 0.15433508789353462,
+      "grad_norm": 0.4104770870479379,
+      "learning_rate": 9.601028263999585e-06,
+      "loss": 0.019,
+      "step": 777
+    },
+    {
+      "epoch": 0.15453371735028304,
+      "grad_norm": 1.0206867249184952,
+      "learning_rate": 9.599767858237976e-06,
+      "loss": 0.0162,
+      "step": 778
+    },
+    {
+      "epoch": 0.1547323468070315,
+      "grad_norm": 0.6045402333325749,
+      "learning_rate": 9.598505547717103e-06,
+      "loss": 0.0171,
+      "step": 779
+    },
+    {
+      "epoch": 0.1549309762637799,
+      "grad_norm": 0.2885932547248781,
+      "learning_rate": 9.597241332959687e-06,
+      "loss": 0.0093,
+      "step": 780
+    },
+    {
+      "epoch": 0.15512960572052836,
+      "grad_norm": 1.0510497521650195,
+      "learning_rate": 9.59597521448924e-06,
+      "loss": 0.0136,
+      "step": 781
+    },
+    {
+      "epoch": 0.15532823517727679,
+      "grad_norm": 0.4857875665691372,
+      "learning_rate": 9.59470719283006e-06,
+      "loss": 0.0151,
+      "step": 782
+    },
+    {
+      "epoch": 0.15552686463402524,
+      "grad_norm": 0.21785943423231574,
+      "learning_rate": 9.59343726850723e-06,
+      "loss": 0.0094,
+      "step": 783
+    },
+    {
+      "epoch": 0.15572549409077366,
+      "grad_norm": 0.365806832334753,
+      "learning_rate": 9.592165442046628e-06,
+      "loss": 0.0141,
+      "step": 784
+    },
+    {
+      "epoch": 0.1559241235475221,
+      "grad_norm": 0.5929826587521397,
+      "learning_rate": 9.590891713974917e-06,
+      "loss": 0.0129,
+      "step": 785
+    },
+    {
+      "epoch": 0.15612275300427053,
+      "grad_norm": 0.7523823133800354,
+      "learning_rate": 9.589616084819542e-06,
+      "loss": 0.0219,
+      "step": 786
+    },
+    {
+      "epoch": 0.15632138246101898,
+      "grad_norm": 0.387882993262946,
+      "learning_rate": 9.588338555108744e-06,
+      "loss": 0.0209,
+      "step": 787
+    },
+    {
+      "epoch": 0.1565200119177674,
+      "grad_norm": 0.45441023735943464,
+      "learning_rate": 9.587059125371545e-06,
+      "loss": 0.0137,
+      "step": 788
+    },
+    {
+      "epoch": 0.15671864137451583,
+      "grad_norm": 0.7600639793039875,
+      "learning_rate": 9.585777796137756e-06,
+      "loss": 0.0114,
+      "step": 789
+    },
+    {
+      "epoch": 0.15691727083126428,
+      "grad_norm": 0.6247443320410846,
+      "learning_rate": 9.584494567937973e-06,
+      "loss": 0.0196,
+      "step": 790
+    },
+    {
+      "epoch": 0.1571159002880127,
+      "grad_norm": 0.2926748180028533,
+      "learning_rate": 9.58320944130358e-06,
+      "loss": 0.0067,
+      "step": 791
+    },
+    {
+      "epoch": 0.15731452974476115,
+      "grad_norm": 0.9470618869437955,
+      "learning_rate": 9.581922416766748e-06,
+      "loss": 0.0207,
+      "step": 792
+    },
+    {
+      "epoch": 0.15751315920150957,
+      "grad_norm": 0.7327000387245999,
+      "learning_rate": 9.580633494860432e-06,
+      "loss": 0.0185,
+      "step": 793
+    },
+    {
+      "epoch": 0.15771178865825802,
+      "grad_norm": 0.5905549844276268,
+      "learning_rate": 9.579342676118373e-06,
+      "loss": 0.015,
+      "step": 794
+    },
+    {
+      "epoch": 0.15791041811500645,
+      "grad_norm": 0.2330283466633709,
+      "learning_rate": 9.578049961075098e-06,
+      "loss": 0.0101,
+      "step": 795
+    },
+    {
+      "epoch": 0.1581090475717549,
+      "grad_norm": 1.4809497082727017,
+      "learning_rate": 9.576755350265918e-06,
+      "loss": 0.0228,
+      "step": 796
+    },
+    {
+      "epoch": 0.15830767702850332,
+      "grad_norm": 0.3008493915594875,
+      "learning_rate": 9.57545884422693e-06,
+      "loss": 0.0078,
+      "step": 797
+    },
+    {
+      "epoch": 0.15850630648525177,
+      "grad_norm": 0.7392324826529579,
+      "learning_rate": 9.574160443495017e-06,
+      "loss": 0.0201,
+      "step": 798
+    },
+    {
+      "epoch": 0.1587049359420002,
+      "grad_norm": 0.546883075423793,
+      "learning_rate": 9.572860148607846e-06,
+      "loss": 0.0173,
+      "step": 799
+    },
+    {
+      "epoch": 0.15890356539874864,
+      "grad_norm": 0.8293724216580041,
+      "learning_rate": 9.571557960103867e-06,
+      "loss": 0.0139,
+      "step": 800
+    },
+    {
+      "epoch": 0.15910219485549706,
+      "grad_norm": 0.9771544832714681,
+      "learning_rate": 9.570253878522314e-06,
+      "loss": 0.0123,
+      "step": 801
+    },
+    {
+      "epoch": 0.15930082431224551,
+      "grad_norm": 0.5950320778242356,
+      "learning_rate": 9.568947904403208e-06,
+      "loss": 0.0177,
+      "step": 802
+    },
+    {
+      "epoch": 0.15949945376899394,
+      "grad_norm": 1.2291334278846218,
+      "learning_rate": 9.567640038287349e-06,
+      "loss": 0.0215,
+      "step": 803
+    },
+    {
+      "epoch": 0.1596980832257424,
+      "grad_norm": 1.1112781913894665,
+      "learning_rate": 9.566330280716323e-06,
+      "loss": 0.0171,
+      "step": 804
+    },
+    {
+      "epoch": 0.1598967126824908,
+      "grad_norm": 0.8798985320697517,
+      "learning_rate": 9.565018632232502e-06,
+      "loss": 0.0245,
+      "step": 805
+    },
+    {
+      "epoch": 0.16009534213923926,
+      "grad_norm": 0.5515295292743605,
+      "learning_rate": 9.563705093379036e-06,
+      "loss": 0.0245,
+      "step": 806
+    },
+    {
+      "epoch": 0.16029397159598768,
+      "grad_norm": 0.6661943953386937,
+      "learning_rate": 9.562389664699863e-06,
+      "loss": 0.0159,
+      "step": 807
+    },
+    {
+      "epoch": 0.16049260105273613,
+      "grad_norm": 0.9464937907087646,
+      "learning_rate": 9.561072346739697e-06,
+      "loss": 0.0148,
+      "step": 808
+    },
+    {
+      "epoch": 0.16069123050948456,
+      "grad_norm": 0.39866831976662964,
+      "learning_rate": 9.55975314004404e-06,
+      "loss": 0.0177,
+      "step": 809
+    },
+    {
+      "epoch": 0.16088985996623298,
+      "grad_norm": 0.6113943733065699,
+      "learning_rate": 9.558432045159174e-06,
+      "loss": 0.0147,
+      "step": 810
+    },
+    {
+      "epoch": 0.16108848942298143,
+      "grad_norm": 0.6538607257959484,
+      "learning_rate": 9.557109062632164e-06,
+      "loss": 0.0226,
+      "step": 811
+    },
+    {
+      "epoch": 0.16128711887972985,
+      "grad_norm": 0.2911939937872243,
+      "learning_rate": 9.555784193010854e-06,
+      "loss": 0.0093,
+      "step": 812
+    },
+    {
+      "epoch": 0.1614857483364783,
+      "grad_norm": 0.43968491191440634,
+      "learning_rate": 9.554457436843872e-06,
+      "loss": 0.0148,
+      "step": 813
+    },
+    {
+      "epoch": 0.16168437779322672,
+      "grad_norm": 1.1183514039831923,
+      "learning_rate": 9.553128794680626e-06,
+      "loss": 0.017,
+      "step": 814
+    },
+    {
+      "epoch": 0.16188300724997517,
+      "grad_norm": 0.6035034096507715,
+      "learning_rate": 9.551798267071308e-06,
+      "loss": 0.0069,
+      "step": 815
+    },
+    {
+      "epoch": 0.1620816367067236,
+      "grad_norm": 0.6846614549208714,
+      "learning_rate": 9.550465854566884e-06,
+      "loss": 0.013,
+      "step": 816
+    },
+    {
+      "epoch": 0.16228026616347205,
+      "grad_norm": 0.5513653606442389,
+      "learning_rate": 9.549131557719106e-06,
+      "loss": 0.0137,
+      "step": 817
+    },
+    {
+      "epoch": 0.16247889562022047,
+      "grad_norm": 0.7496807116130924,
+      "learning_rate": 9.547795377080506e-06,
+      "loss": 0.0062,
+      "step": 818
+    },
+    {
+      "epoch": 0.16267752507696892,
+      "grad_norm": 1.1195510094425656,
+      "learning_rate": 9.546457313204395e-06,
+      "loss": 0.0186,
+      "step": 819
+    },
+    {
+      "epoch": 0.16287615453371734,
+      "grad_norm": 0.4918213514406738,
+      "learning_rate": 9.545117366644863e-06,
+      "loss": 0.0082,
+      "step": 820
+    },
+    {
+      "epoch": 0.1630747839904658,
+      "grad_norm": 0.6864874480930118,
+      "learning_rate": 9.543775537956781e-06,
+      "loss": 0.0165,
+      "step": 821
+    },
+    {
+      "epoch": 0.16327341344721422,
+      "grad_norm": 0.9104337265905692,
+      "learning_rate": 9.5424318276958e-06,
+      "loss": 0.0202,
+      "step": 822
+    },
+    {
+      "epoch": 0.16347204290396267,
+      "grad_norm": 1.3936891736313024,
+      "learning_rate": 9.541086236418348e-06,
+      "loss": 0.0139,
+      "step": 823
+    },
+    {
+      "epoch": 0.1636706723607111,
+      "grad_norm": 1.0710993378946527,
+      "learning_rate": 9.539738764681633e-06,
+      "loss": 0.0207,
+      "step": 824
+    },
+    {
+      "epoch": 0.16386930181745954,
+      "grad_norm": 0.7599724122468066,
+      "learning_rate": 9.538389413043641e-06,
+      "loss": 0.0194,
+      "step": 825
+    },
+    {
+      "epoch": 0.16406793127420796,
+      "grad_norm": 0.4138470932916712,
+      "learning_rate": 9.537038182063138e-06,
+      "loss": 0.009,
+      "step": 826
+    },
+    {
+      "epoch": 0.1642665607309564,
+      "grad_norm": 1.0081381826347446,
+      "learning_rate": 9.535685072299668e-06,
+      "loss": 0.0172,
+      "step": 827
+    },
+    {
+      "epoch": 0.16446519018770484,
+      "grad_norm": 1.0432367672352847,
+      "learning_rate": 9.53433008431355e-06,
+      "loss": 0.0161,
+      "step": 828
+    },
+    {
+      "epoch": 0.16466381964445329,
+      "grad_norm": 0.7812431690369702,
+      "learning_rate": 9.532973218665887e-06,
+      "loss": 0.0237,
+      "step": 829
+    },
+    {
+      "epoch": 0.1648624491012017,
+      "grad_norm": 0.4534298510315967,
+      "learning_rate": 9.531614475918552e-06,
+      "loss": 0.0138,
+      "step": 830
+    },
+    {
+      "epoch": 0.16506107855795013,
+      "grad_norm": 0.521000256028603,
+      "learning_rate": 9.530253856634202e-06,
+      "loss": 0.014,
+      "step": 831
+    },
+    {
+      "epoch": 0.16525970801469858,
+      "grad_norm": 0.4111263927903728,
+      "learning_rate": 9.528891361376265e-06,
+      "loss": 0.0102,
+      "step": 832
+    },
+    {
+      "epoch": 0.165458337471447,
+      "grad_norm": 0.6400161006473448,
+      "learning_rate": 9.527526990708952e-06,
+      "loss": 0.0135,
+      "step": 833
+    },
+    {
+      "epoch": 0.16565696692819545,
+      "grad_norm": 1.0992489588141927,
+      "learning_rate": 9.526160745197247e-06,
+      "loss": 0.022,
+      "step": 834
+    },
+    {
+      "epoch": 0.16585559638494388,
+      "grad_norm": 0.2568053298935382,
+      "learning_rate": 9.524792625406908e-06,
+      "loss": 0.0133,
+      "step": 835
+    },
+    {
+      "epoch": 0.16605422584169233,
+      "grad_norm": 0.6766945867655767,
+      "learning_rate": 9.523422631904473e-06,
+      "loss": 0.0104,
+      "step": 836
+    },
+    {
+      "epoch": 0.16625285529844075,
+      "grad_norm": 0.5619582062884672,
+      "learning_rate": 9.522050765257257e-06,
+      "loss": 0.0253,
+      "step": 837
+    },
+    {
+      "epoch": 0.1664514847551892,
+      "grad_norm": 0.4221715846394782,
+      "learning_rate": 9.52067702603335e-06,
+      "loss": 0.0151,
+      "step": 838
+    },
+    {
+      "epoch": 0.16665011421193762,
+      "grad_norm": 0.22009747488862705,
+      "learning_rate": 9.519301414801612e-06,
+      "loss": 0.0089,
+      "step": 839
+    },
+    {
+      "epoch": 0.16684874366868607,
+      "grad_norm": 0.6287660735362872,
+      "learning_rate": 9.517923932131685e-06,
+      "loss": 0.0216,
+      "step": 840
+    },
+    {
+      "epoch": 0.1670473731254345,
+      "grad_norm": 0.4170499590382144,
+      "learning_rate": 9.516544578593981e-06,
+      "loss": 0.0108,
+      "step": 841
+    },
+    {
+      "epoch": 0.16724600258218295,
+      "grad_norm": 0.9421745387040598,
+      "learning_rate": 9.51516335475969e-06,
+      "loss": 0.0177,
+      "step": 842
+    },
+    {
+      "epoch": 0.16744463203893137,
+      "grad_norm": 0.4757860318481506,
+      "learning_rate": 9.513780261200774e-06,
+      "loss": 0.011,
+      "step": 843
+    },
+    {
+      "epoch": 0.16764326149567982,
+      "grad_norm": 0.7429722294449872,
+      "learning_rate": 9.512395298489974e-06,
+      "loss": 0.0148,
+      "step": 844
+    },
+    {
+      "epoch": 0.16784189095242824,
+      "grad_norm": 0.5206174876646718,
+      "learning_rate": 9.511008467200798e-06,
+      "loss": 0.0156,
+      "step": 845
+    },
+    {
+      "epoch": 0.1680405204091767,
+      "grad_norm": 1.3998688162229813,
+      "learning_rate": 9.509619767907534e-06,
+      "loss": 0.0239,
+      "step": 846
+    },
+    {
+      "epoch": 0.16823914986592511,
+      "grad_norm": 0.8208188686925355,
+      "learning_rate": 9.508229201185242e-06,
+      "loss": 0.0135,
+      "step": 847
+    },
+    {
+      "epoch": 0.16843777932267356,
+      "grad_norm": 0.3774258541529112,
+      "learning_rate": 9.506836767609751e-06,
+      "loss": 0.014,
+      "step": 848
+    },
+    {
+      "epoch": 0.168636408779422,
+      "grad_norm": 0.668337123256141,
+      "learning_rate": 9.505442467757666e-06,
+      "loss": 0.013,
+      "step": 849
+    },
+    {
+      "epoch": 0.16883503823617044,
+      "grad_norm": 0.4690654187501145,
+      "learning_rate": 9.504046302206368e-06,
+      "loss": 0.0229,
+      "step": 850
+    },
+    {
+      "epoch": 0.16903366769291886,
+      "grad_norm": 0.9382417671995416,
+      "learning_rate": 9.50264827153401e-06,
+      "loss": 0.0201,
+      "step": 851
+    },
+    {
+      "epoch": 0.16923229714966728,
+      "grad_norm": 0.5705256377710812,
+      "learning_rate": 9.501248376319508e-06,
+      "loss": 0.0128,
+      "step": 852
+    },
+    {
+      "epoch": 0.16943092660641573,
+      "grad_norm": 0.48645671259102613,
+      "learning_rate": 9.499846617142563e-06,
+      "loss": 0.014,
+      "step": 853
+    },
+    {
+      "epoch": 0.16962955606316416,
+      "grad_norm": 1.4022653145077122,
+      "learning_rate": 9.498442994583639e-06,
+      "loss": 0.0174,
+      "step": 854
+    },
+    {
+      "epoch": 0.1698281855199126,
+      "grad_norm": 0.7638533431862135,
+      "learning_rate": 9.497037509223977e-06,
+      "loss": 0.0149,
+      "step": 855
+    },
+    {
+      "epoch": 0.17002681497666103,
+      "grad_norm": 1.519768782139612,
+      "learning_rate": 9.495630161645584e-06,
+      "loss": 0.0255,
+      "step": 856
+    },
+    {
+      "epoch": 0.17022544443340948,
+      "grad_norm": 0.44744869619282945,
+      "learning_rate": 9.494220952431243e-06,
+      "loss": 0.0132,
+      "step": 857
+    },
+    {
+      "epoch": 0.1704240738901579,
+      "grad_norm": 0.5073442816162103,
+      "learning_rate": 9.492809882164509e-06,
+      "loss": 0.0219,
+      "step": 858
+    },
+    {
+      "epoch": 0.17062270334690635,
+      "grad_norm": 0.6578886831791138,
+      "learning_rate": 9.491396951429698e-06,
+      "loss": 0.0118,
+      "step": 859
+    },
+    {
+      "epoch": 0.17082133280365477,
+      "grad_norm": 0.440174527734574,
+      "learning_rate": 9.48998216081191e-06,
+      "loss": 0.023,
+      "step": 860
+    },
+    {
+      "epoch": 0.17101996226040322,
+      "grad_norm": 1.2110068679815877,
+      "learning_rate": 9.488565510897006e-06,
+      "loss": 0.0213,
+      "step": 861
+    },
+    {
+      "epoch": 0.17121859171715165,
+      "grad_norm": 0.34225622494013835,
+      "learning_rate": 9.487147002271618e-06,
+      "loss": 0.0218,
+      "step": 862
+    },
+    {
+      "epoch": 0.1714172211739001,
+      "grad_norm": 0.9248969475348255,
+      "learning_rate": 9.48572663552315e-06,
+      "loss": 0.0172,
+      "step": 863
+    },
+    {
+      "epoch": 0.17161585063064852,
+      "grad_norm": 0.3111389224565366,
+      "learning_rate": 9.484304411239774e-06,
+      "loss": 0.0194,
+      "step": 864
+    },
+    {
+      "epoch": 0.17181448008739697,
+      "grad_norm": 0.5426022611346796,
+      "learning_rate": 9.482880330010434e-06,
+      "loss": 0.0177,
+      "step": 865
+    },
+    {
+      "epoch": 0.1720131095441454,
+      "grad_norm": 0.9989174638535595,
+      "learning_rate": 9.481454392424836e-06,
+      "loss": 0.0179,
+      "step": 866
+    },
+    {
+      "epoch": 0.17221173900089384,
+      "grad_norm": 1.0032440886756715,
+      "learning_rate": 9.480026599073463e-06,
+      "loss": 0.0214,
+      "step": 867
+    },
+    {
+      "epoch": 0.17241036845764227,
+      "grad_norm": 0.29351054118887016,
+      "learning_rate": 9.478596950547561e-06,
+      "loss": 0.0084,
+      "step": 868
+    },
+    {
+      "epoch": 0.17260899791439072,
+      "grad_norm": 0.5469059731359652,
+      "learning_rate": 9.477165447439148e-06,
+      "loss": 0.0228,
+      "step": 869
+    },
+    {
+      "epoch": 0.17280762737113914,
+      "grad_norm": 0.5151774165372478,
+      "learning_rate": 9.475732090341006e-06,
+      "loss": 0.017,
+      "step": 870
+    },
+    {
+      "epoch": 0.1730062568278876,
+      "grad_norm": 0.3849241059477399,
+      "learning_rate": 9.474296879846688e-06,
+      "loss": 0.0149,
+      "step": 871
+    },
+    {
+      "epoch": 0.173204886284636,
+      "grad_norm": 0.2846207510615562,
+      "learning_rate": 9.47285981655051e-06,
+      "loss": 0.0134,
+      "step": 872
+    },
+    {
+      "epoch": 0.17340351574138443,
+      "grad_norm": 0.3822669264577504,
+      "learning_rate": 9.471420901047564e-06,
+      "loss": 0.0126,
+      "step": 873
+    },
+    {
+      "epoch": 0.17360214519813288,
+      "grad_norm": 0.5809048365250173,
+      "learning_rate": 9.469980133933701e-06,
+      "loss": 0.0186,
+      "step": 874
+    },
+    {
+      "epoch": 0.1738007746548813,
+      "grad_norm": 0.8463872528550324,
+      "learning_rate": 9.46853751580554e-06,
+      "loss": 0.0195,
+      "step": 875
+    },
+    {
+      "epoch": 0.17399940411162976,
+      "grad_norm": 0.4607531534725536,
+      "learning_rate": 9.467093047260468e-06,
+      "loss": 0.0151,
+      "step": 876
+    },
+    {
+      "epoch": 0.17419803356837818,
+      "grad_norm": 1.1647174596925964,
+      "learning_rate": 9.465646728896641e-06,
+      "loss": 0.0138,
+      "step": 877
+    },
+    {
+      "epoch": 0.17439666302512663,
+      "grad_norm": 0.6812902706877884,
+      "learning_rate": 9.464198561312972e-06,
+      "loss": 0.0146,
+      "step": 878
+    },
+    {
+      "epoch": 0.17459529248187505,
+      "grad_norm": 1.118536537307821,
+      "learning_rate": 9.462748545109152e-06,
+      "loss": 0.0144,
+      "step": 879
+    },
+    {
+      "epoch": 0.1747939219386235,
+      "grad_norm": 0.6063904311802616,
+      "learning_rate": 9.461296680885628e-06,
+      "loss": 0.0073,
+      "step": 880
+    },
+    {
+      "epoch": 0.17499255139537193,
+      "grad_norm": 1.5335959139523947,
+      "learning_rate": 9.459842969243615e-06,
+      "loss": 0.0237,
+      "step": 881
+    },
+    {
+      "epoch": 0.17519118085212038,
+      "grad_norm": 0.4453751489154979,
+      "learning_rate": 9.458387410785096e-06,
+      "loss": 0.0129,
+      "step": 882
+    },
+    {
+      "epoch": 0.1753898103088688,
+      "grad_norm": 0.5312389194281991,
+      "learning_rate": 9.456930006112814e-06,
+      "loss": 0.0231,
+      "step": 883
+    },
+    {
+      "epoch": 0.17558843976561725,
+      "grad_norm": 0.5246562462701451,
+      "learning_rate": 9.45547075583028e-06,
+      "loss": 0.0141,
+      "step": 884
+    },
+    {
+      "epoch": 0.17578706922236567,
+      "grad_norm": 0.3711081470488199,
+      "learning_rate": 9.454009660541769e-06,
+      "loss": 0.0126,
+      "step": 885
+    },
+    {
+      "epoch": 0.17598569867911412,
+      "grad_norm": 0.4214526210893554,
+      "learning_rate": 9.452546720852317e-06,
+      "loss": 0.0098,
+      "step": 886
+    },
+    {
+      "epoch": 0.17618432813586254,
+      "grad_norm": 0.730638217900042,
+      "learning_rate": 9.451081937367725e-06,
+      "loss": 0.0155,
+      "step": 887
+    },
+    {
+      "epoch": 0.176382957592611,
+      "grad_norm": 0.5304446182299737,
+      "learning_rate": 9.449615310694563e-06,
+      "loss": 0.0158,
+      "step": 888
+    },
+    {
+      "epoch": 0.17658158704935942,
+      "grad_norm": 0.4787768591240991,
+      "learning_rate": 9.448146841440156e-06,
+      "loss": 0.011,
+      "step": 889
+    },
+    {
+      "epoch": 0.17678021650610787,
+      "grad_norm": 0.6573768696874656,
+      "learning_rate": 9.446676530212596e-06,
+      "loss": 0.0121,
+      "step": 890
+    },
+    {
+      "epoch": 0.1769788459628563,
+      "grad_norm": 0.8738613272343506,
+      "learning_rate": 9.445204377620739e-06,
+      "loss": 0.0218,
+      "step": 891
+    },
+    {
+      "epoch": 0.17717747541960474,
+      "grad_norm": 0.6468029328396412,
+      "learning_rate": 9.443730384274199e-06,
+      "loss": 0.0126,
+      "step": 892
+    },
+    {
+      "epoch": 0.17737610487635316,
+      "grad_norm": 0.6916082592292682,
+      "learning_rate": 9.442254550783357e-06,
+      "loss": 0.0165,
+      "step": 893
+    },
+    {
+      "epoch": 0.17757473433310159,
+      "grad_norm": 0.4406671563052852,
+      "learning_rate": 9.440776877759354e-06,
+      "loss": 0.0115,
+      "step": 894
+    },
+    {
+      "epoch": 0.17777336378985004,
+      "grad_norm": 1.0537465676266842,
+      "learning_rate": 9.439297365814095e-06,
+      "loss": 0.0174,
+      "step": 895
+    },
+    {
+      "epoch": 0.17797199324659846,
+      "grad_norm": 1.2213613739765894,
+      "learning_rate": 9.437816015560241e-06,
+      "loss": 0.0183,
+      "step": 896
+    },
+    {
+      "epoch": 0.1781706227033469,
+      "grad_norm": 0.4783500290161957,
+      "learning_rate": 9.43633282761122e-06,
+      "loss": 0.0092,
+      "step": 897
+    },
+    {
+      "epoch": 0.17836925216009533,
+      "grad_norm": 0.43432963042155853,
+      "learning_rate": 9.434847802581216e-06,
+      "loss": 0.0203,
+      "step": 898
+    },
+    {
+      "epoch": 0.17856788161684378,
+      "grad_norm": 0.7897317815340374,
+      "learning_rate": 9.43336094108518e-06,
+      "loss": 0.0108,
+      "step": 899
+    },
+    {
+      "epoch": 0.1787665110735922,
+      "grad_norm": 0.3008681013432827,
+      "learning_rate": 9.431872243738817e-06,
+      "loss": 0.0137,
+      "step": 900
+    },
+    {
+      "epoch": 0.17896514053034066,
+      "grad_norm": 0.6771184704989937,
+      "learning_rate": 9.430381711158597e-06,
+      "loss": 0.0262,
+      "step": 901
+    },
+    {
+      "epoch": 0.17916376998708908,
+      "grad_norm": 0.4029134503869768,
+      "learning_rate": 9.428889343961745e-06,
+      "loss": 0.0082,
+      "step": 902
+    },
+    {
+      "epoch": 0.17936239944383753,
+      "grad_norm": 0.6499915119874066,
+      "learning_rate": 9.427395142766253e-06,
+      "loss": 0.0104,
+      "step": 903
+    },
+    {
+      "epoch": 0.17956102890058595,
+      "grad_norm": 1.2873816596565066,
+      "learning_rate": 9.425899108190866e-06,
+      "loss": 0.014,
+      "step": 904
+    },
+    {
+      "epoch": 0.1797596583573344,
+      "grad_norm": 0.917742932464249,
+      "learning_rate": 9.42440124085509e-06,
+      "loss": 0.0098,
+      "step": 905
+    },
+    {
+      "epoch": 0.17995828781408282,
+      "grad_norm": 0.6917074788463903,
+      "learning_rate": 9.42290154137919e-06,
+      "loss": 0.0111,
+      "step": 906
+    },
+    {
+      "epoch": 0.18015691727083127,
+      "grad_norm": 0.4982836344620947,
+      "learning_rate": 9.421400010384191e-06,
+      "loss": 0.0145,
+      "step": 907
+    },
+    {
+      "epoch": 0.1803555467275797,
+      "grad_norm": 1.086400992320905,
+      "learning_rate": 9.419896648491875e-06,
+      "loss": 0.0174,
+      "step": 908
+    },
+    {
+      "epoch": 0.18055417618432815,
+      "grad_norm": 0.8193422731728996,
+      "learning_rate": 9.418391456324785e-06,
+      "loss": 0.0105,
+      "step": 909
+    },
+    {
+      "epoch": 0.18075280564107657,
+      "grad_norm": 0.7103752650742982,
+      "learning_rate": 9.416884434506217e-06,
+      "loss": 0.0139,
+      "step": 910
+    },
+    {
+      "epoch": 0.18095143509782502,
+      "grad_norm": 0.6962334695928468,
+      "learning_rate": 9.415375583660227e-06,
+      "loss": 0.0128,
+      "step": 911
+    },
+    {
+      "epoch": 0.18115006455457344,
+      "grad_norm": 0.6806563393000292,
+      "learning_rate": 9.41386490441163e-06,
+      "loss": 0.0151,
+      "step": 912
+    },
+    {
+      "epoch": 0.1813486940113219,
+      "grad_norm": 0.477963378101734,
+      "learning_rate": 9.412352397385997e-06,
+      "loss": 0.0156,
+      "step": 913
+    },
+    {
+      "epoch": 0.18154732346807032,
+      "grad_norm": 1.377024847688744,
+      "learning_rate": 9.410838063209653e-06,
+      "loss": 0.0137,
+      "step": 914
+    },
+    {
+      "epoch": 0.18174595292481874,
+      "grad_norm": 0.840278914999452,
+      "learning_rate": 9.409321902509686e-06,
+      "loss": 0.0179,
+      "step": 915
+    },
+    {
+      "epoch": 0.1819445823815672,
+      "grad_norm": 1.2080657080079076,
+      "learning_rate": 9.407803915913934e-06,
+      "loss": 0.0142,
+      "step": 916
+    },
+    {
+      "epoch": 0.1821432118383156,
+      "grad_norm": 0.6470778636335309,
+      "learning_rate": 9.406284104050994e-06,
+      "loss": 0.0153,
+      "step": 917
+    },
+    {
+      "epoch": 0.18234184129506406,
+      "grad_norm": 0.6448817205204879,
+      "learning_rate": 9.40476246755022e-06,
+      "loss": 0.0147,
+      "step": 918
+    },
+    {
+      "epoch": 0.18254047075181248,
+      "grad_norm": 0.8677307803730453,
+      "learning_rate": 9.403239007041719e-06,
+      "loss": 0.0224,
+      "step": 919
+    },
+    {
+      "epoch": 0.18273910020856093,
+      "grad_norm": 1.0883235135924039,
+      "learning_rate": 9.401713723156355e-06,
+      "loss": 0.0156,
+      "step": 920
+    },
+    {
+      "epoch": 0.18293772966530936,
+      "grad_norm": 1.378109947347798,
+      "learning_rate": 9.400186616525747e-06,
+      "loss": 0.0214,
+      "step": 921
+    },
+    {
+      "epoch": 0.1831363591220578,
+      "grad_norm": 0.45828575967420676,
+      "learning_rate": 9.398657687782264e-06,
+      "loss": 0.0143,
+      "step": 922
+    },
+    {
+      "epoch": 0.18333498857880623,
+      "grad_norm": 0.4067470046392445,
+      "learning_rate": 9.397126937559041e-06,
+      "loss": 0.0106,
+      "step": 923
+    },
+    {
+      "epoch": 0.18353361803555468,
+      "grad_norm": 0.8655229021701957,
+      "learning_rate": 9.395594366489956e-06,
+      "loss": 0.0164,
+      "step": 924
+    },
+    {
+      "epoch": 0.1837322474923031,
+      "grad_norm": 0.6544805313594291,
+      "learning_rate": 9.394059975209644e-06,
+      "loss": 0.0161,
+      "step": 925
+    },
+    {
+      "epoch": 0.18393087694905155,
+      "grad_norm": 0.5260990122458291,
+      "learning_rate": 9.392523764353497e-06,
+      "loss": 0.0121,
+      "step": 926
+    },
+    {
+      "epoch": 0.18412950640579998,
+      "grad_norm": 0.49273647924970304,
+      "learning_rate": 9.390985734557659e-06,
+      "loss": 0.0137,
+      "step": 927
+    },
+    {
+      "epoch": 0.18432813586254843,
+      "grad_norm": 0.5039646852586156,
+      "learning_rate": 9.389445886459026e-06,
+      "loss": 0.0118,
+      "step": 928
+    },
+    {
+      "epoch": 0.18452676531929685,
+      "grad_norm": 0.6159677479137697,
+      "learning_rate": 9.387904220695245e-06,
+      "loss": 0.0096,
+      "step": 929
+    },
+    {
+      "epoch": 0.1847253947760453,
+      "grad_norm": 0.4398053369763513,
+      "learning_rate": 9.386360737904722e-06,
+      "loss": 0.0082,
+      "step": 930
+    },
+    {
+      "epoch": 0.18492402423279372,
+      "grad_norm": 0.5382584798731853,
+      "learning_rate": 9.384815438726608e-06,
+      "loss": 0.0143,
+      "step": 931
+    },
+    {
+      "epoch": 0.18512265368954217,
+      "grad_norm": 0.7022731175340937,
+      "learning_rate": 9.383268323800815e-06,
+      "loss": 0.0231,
+      "step": 932
+    },
+    {
+      "epoch": 0.1853212831462906,
+      "grad_norm": 0.45625159282982203,
+      "learning_rate": 9.381719393767998e-06,
+      "loss": 0.018,
+      "step": 933
+    },
+    {
+      "epoch": 0.18551991260303904,
+      "grad_norm": 0.5960828207873747,
+      "learning_rate": 9.380168649269566e-06,
+      "loss": 0.0147,
+      "step": 934
+    },
+    {
+      "epoch": 0.18571854205978747,
+      "grad_norm": 0.8282241450571749,
+      "learning_rate": 9.378616090947685e-06,
+      "loss": 0.0215,
+      "step": 935
+    },
+    {
+      "epoch": 0.1859171715165359,
+      "grad_norm": 0.4426456726841343,
+      "learning_rate": 9.377061719445264e-06,
+      "loss": 0.0229,
+      "step": 936
+    },
+    {
+      "epoch": 0.18611580097328434,
+      "grad_norm": 0.5920397094573076,
+      "learning_rate": 9.375505535405969e-06,
+      "loss": 0.0173,
+      "step": 937
+    },
+    {
+      "epoch": 0.18631443043003276,
+      "grad_norm": 0.2981282792836881,
+      "learning_rate": 9.373947539474212e-06,
+      "loss": 0.0112,
+      "step": 938
+    },
+    {
+      "epoch": 0.1865130598867812,
+      "grad_norm": 0.7856812810221333,
+      "learning_rate": 9.372387732295162e-06,
+      "loss": 0.0119,
+      "step": 939
+    },
+    {
+      "epoch": 0.18671168934352964,
+      "grad_norm": 0.7020478883034995,
+      "learning_rate": 9.370826114514729e-06,
+      "loss": 0.0138,
+      "step": 940
+    },
+    {
+      "epoch": 0.18691031880027809,
+      "grad_norm": 0.3879630472447985,
+      "learning_rate": 9.369262686779578e-06,
+      "loss": 0.0098,
+      "step": 941
+    },
+    {
+      "epoch": 0.1871089482570265,
+      "grad_norm": 0.26290487954249336,
+      "learning_rate": 9.367697449737126e-06,
+      "loss": 0.0086,
+      "step": 942
+    },
+    {
+      "epoch": 0.18730757771377496,
+      "grad_norm": 0.8158600730845823,
+      "learning_rate": 9.366130404035533e-06,
+      "loss": 0.0127,
+      "step": 943
+    },
+    {
+      "epoch": 0.18750620717052338,
+      "grad_norm": 0.43011489476490694,
+      "learning_rate": 9.364561550323711e-06,
+      "loss": 0.0138,
+      "step": 944
+    },
+    {
+      "epoch": 0.18770483662727183,
+      "grad_norm": 0.6422387414139117,
+      "learning_rate": 9.362990889251325e-06,
+      "loss": 0.0183,
+      "step": 945
+    },
+    {
+      "epoch": 0.18790346608402025,
+      "grad_norm": 0.8940322891219529,
+      "learning_rate": 9.361418421468777e-06,
+      "loss": 0.0187,
+      "step": 946
+    },
+    {
+      "epoch": 0.1881020955407687,
+      "grad_norm": 1.4463527888545427,
+      "learning_rate": 9.359844147627231e-06,
+      "loss": 0.0214,
+      "step": 947
+    },
+    {
+      "epoch": 0.18830072499751713,
+      "grad_norm": 0.4102746620553028,
+      "learning_rate": 9.358268068378589e-06,
+      "loss": 0.0085,
+      "step": 948
+    },
+    {
+      "epoch": 0.18849935445426558,
+      "grad_norm": 0.6618029175469268,
+      "learning_rate": 9.356690184375504e-06,
+      "loss": 0.0168,
+      "step": 949
+    },
+    {
+      "epoch": 0.188697983911014,
+      "grad_norm": 0.5539961291378328,
+      "learning_rate": 9.355110496271376e-06,
+      "loss": 0.0118,
+      "step": 950
+    },
+    {
+      "epoch": 0.18889661336776245,
+      "grad_norm": 0.8598663158866856,
+      "learning_rate": 9.353529004720354e-06,
+      "loss": 0.0119,
+      "step": 951
+    },
+    {
+      "epoch": 0.18909524282451087,
+      "grad_norm": 0.5890081936780608,
+      "learning_rate": 9.35194571037733e-06,
+      "loss": 0.0199,
+      "step": 952
+    },
+    {
+      "epoch": 0.18929387228125932,
+      "grad_norm": 1.2607952878561444,
+      "learning_rate": 9.350360613897945e-06,
+      "loss": 0.021,
+      "step": 953
+    },
+    {
+      "epoch": 0.18949250173800775,
+      "grad_norm": 0.5805466556115377,
+      "learning_rate": 9.348773715938587e-06,
+      "loss": 0.0103,
+      "step": 954
+    },
+    {
+      "epoch": 0.18969113119475617,
+      "grad_norm": 0.6262277582379767,
+      "learning_rate": 9.347185017156388e-06,
+      "loss": 0.014,
+      "step": 955
+    },
+    {
+      "epoch": 0.18988976065150462,
+      "grad_norm": 0.8635643634403196,
+      "learning_rate": 9.345594518209227e-06,
+      "loss": 0.02,
+      "step": 956
+    },
+    {
+      "epoch": 0.19008839010825304,
+      "grad_norm": 0.4163572607700282,
+      "learning_rate": 9.344002219755728e-06,
+      "loss": 0.0099,
+      "step": 957
+    },
+    {
+      "epoch": 0.1902870195650015,
+      "grad_norm": 0.5499649933024405,
+      "learning_rate": 9.34240812245526e-06,
+      "loss": 0.0123,
+      "step": 958
+    },
+    {
+      "epoch": 0.19048564902174991,
+      "grad_norm": 0.4621935969511572,
+      "learning_rate": 9.340812226967936e-06,
+      "loss": 0.0141,
+      "step": 959
+    },
+    {
+      "epoch": 0.19068427847849836,
+      "grad_norm": 0.6322621882728303,
+      "learning_rate": 9.339214533954618e-06,
+      "loss": 0.0138,
+      "step": 960
+    },
+    {
+      "epoch": 0.1908829079352468,
+      "grad_norm": 0.5315385506545031,
+      "learning_rate": 9.337615044076906e-06,
+      "loss": 0.0156,
+      "step": 961
+    },
+    {
+      "epoch": 0.19108153739199524,
+      "grad_norm": 0.8395315780253293,
+      "learning_rate": 9.336013757997147e-06,
+      "loss": 0.014,
+      "step": 962
+    },
+    {
+      "epoch": 0.19128016684874366,
+      "grad_norm": 0.515182257587864,
+      "learning_rate": 9.334410676378433e-06,
+      "loss": 0.0151,
+      "step": 963
+    },
+    {
+      "epoch": 0.1914787963054921,
+      "grad_norm": 0.7503607628649587,
+      "learning_rate": 9.3328057998846e-06,
+      "loss": 0.0181,
+      "step": 964
+    },
+    {
+      "epoch": 0.19167742576224053,
+      "grad_norm": 0.6292960608616276,
+      "learning_rate": 9.331199129180224e-06,
+      "loss": 0.0256,
+      "step": 965
+    },
+    {
+      "epoch": 0.19187605521898898,
+      "grad_norm": 0.657172738095144,
+      "learning_rate": 9.329590664930625e-06,
+      "loss": 0.0131,
+      "step": 966
+    },
+    {
+      "epoch": 0.1920746846757374,
+      "grad_norm": 0.6758585625636307,
+      "learning_rate": 9.32798040780187e-06,
+      "loss": 0.0233,
+      "step": 967
+    },
+    {
+      "epoch": 0.19227331413248586,
+      "grad_norm": 1.138944554532619,
+      "learning_rate": 9.326368358460757e-06,
+      "loss": 0.0166,
+      "step": 968
+    },
+    {
+      "epoch": 0.19247194358923428,
+      "grad_norm": 0.9844802183689632,
+      "learning_rate": 9.324754517574844e-06,
+      "loss": 0.0147,
+      "step": 969
+    },
+    {
+      "epoch": 0.19267057304598273,
+      "grad_norm": 0.36727739091125694,
+      "learning_rate": 9.323138885812416e-06,
+      "loss": 0.0157,
+      "step": 970
+    },
+    {
+      "epoch": 0.19286920250273115,
+      "grad_norm": 0.39266585464073445,
+      "learning_rate": 9.3215214638425e-06,
+      "loss": 0.0147,
+      "step": 971
+    },
+    {
+      "epoch": 0.1930678319594796,
+      "grad_norm": 0.694569390497426,
+      "learning_rate": 9.319902252334878e-06,
+      "loss": 0.0132,
+      "step": 972
+    },
+    {
+      "epoch": 0.19326646141622802,
+      "grad_norm": 0.8122752981584541,
+      "learning_rate": 9.318281251960059e-06,
+      "loss": 0.0152,
+      "step": 973
+    },
+    {
+      "epoch": 0.19346509087297648,
+      "grad_norm": 1.356751012776408,
+      "learning_rate": 9.316658463389296e-06,
+      "loss": 0.0159,
+      "step": 974
+    },
+    {
+      "epoch": 0.1936637203297249,
+      "grad_norm": 0.4514146325071209,
+      "learning_rate": 9.315033887294588e-06,
+      "loss": 0.0085,
+      "step": 975
+    },
+    {
+      "epoch": 0.19386234978647332,
+      "grad_norm": 0.4492629322519748,
+      "learning_rate": 9.313407524348667e-06,
+      "loss": 0.0121,
+      "step": 976
+    },
+    {
+      "epoch": 0.19406097924322177,
+      "grad_norm": 0.42766136851441033,
+      "learning_rate": 9.311779375225012e-06,
+      "loss": 0.0113,
+      "step": 977
+    },
+    {
+      "epoch": 0.1942596086999702,
+      "grad_norm": 0.3782072201329815,
+      "learning_rate": 9.310149440597833e-06,
+      "loss": 0.006,
+      "step": 978
+    },
+    {
+      "epoch": 0.19445823815671864,
+      "grad_norm": 1.3010134952460966,
+      "learning_rate": 9.308517721142088e-06,
+      "loss": 0.0219,
+      "step": 979
+    },
+    {
+      "epoch": 0.19465686761346707,
+      "grad_norm": 1.7188646382221526,
+      "learning_rate": 9.30688421753347e-06,
+      "loss": 0.021,
+      "step": 980
+    },
+    {
+      "epoch": 0.19485549707021552,
+      "grad_norm": 1.113751111511706,
+      "learning_rate": 9.30524893044841e-06,
+      "loss": 0.0213,
+      "step": 981
+    },
+    {
+      "epoch": 0.19505412652696394,
+      "grad_norm": 0.33969922760265947,
+      "learning_rate": 9.303611860564079e-06,
+      "loss": 0.0088,
+      "step": 982
+    },
+    {
+      "epoch": 0.1952527559837124,
+      "grad_norm": 0.6797604296474096,
+      "learning_rate": 9.301973008558387e-06,
+      "loss": 0.0187,
+      "step": 983
+    },
+    {
+      "epoch": 0.1954513854404608,
+      "grad_norm": 0.7594684777223187,
+      "learning_rate": 9.30033237510998e-06,
+      "loss": 0.0195,
+      "step": 984
+    },
+    {
+      "epoch": 0.19565001489720926,
+      "grad_norm": 1.3637727775160045,
+      "learning_rate": 9.298689960898242e-06,
+      "loss": 0.0126,
+      "step": 985
+    },
+    {
+      "epoch": 0.19584864435395768,
+      "grad_norm": 1.0630474301871424,
+      "learning_rate": 9.297045766603297e-06,
+      "loss": 0.0141,
+      "step": 986
+    },
+    {
+      "epoch": 0.19604727381070614,
+      "grad_norm": 0.6542337994520813,
+      "learning_rate": 9.295399792906002e-06,
+      "loss": 0.021,
+      "step": 987
+    },
+    {
+      "epoch": 0.19624590326745456,
+      "grad_norm": 0.5665306835890809,
+      "learning_rate": 9.293752040487956e-06,
+      "loss": 0.0142,
+      "step": 988
+    },
+    {
+      "epoch": 0.196444532724203,
+      "grad_norm": 0.7553552246755394,
+      "learning_rate": 9.292102510031488e-06,
+      "loss": 0.0132,
+      "step": 989
+    },
+    {
+      "epoch": 0.19664316218095143,
+      "grad_norm": 1.470905402981988,
+      "learning_rate": 9.29045120221967e-06,
+      "loss": 0.0256,
+      "step": 990
+    },
+    {
+      "epoch": 0.19684179163769988,
+      "grad_norm": 0.5202547467174007,
+      "learning_rate": 9.288798117736307e-06,
+      "loss": 0.0128,
+      "step": 991
+    },
+    {
+      "epoch": 0.1970404210944483,
+      "grad_norm": 0.48524424654365406,
+      "learning_rate": 9.287143257265936e-06,
+      "loss": 0.0135,
+      "step": 992
+    },
+    {
+      "epoch": 0.19723905055119675,
+      "grad_norm": 0.6718955314932935,
+      "learning_rate": 9.285486621493836e-06,
+      "loss": 0.0125,
+      "step": 993
+    },
+    {
+      "epoch": 0.19743768000794518,
+      "grad_norm": 0.8034984592157217,
+      "learning_rate": 9.283828211106019e-06,
+      "loss": 0.0199,
+      "step": 994
+    },
+    {
+      "epoch": 0.19763630946469363,
+      "grad_norm": 0.5390976895123648,
+      "learning_rate": 9.28216802678923e-06,
+      "loss": 0.0131,
+      "step": 995
+    },
+    {
+      "epoch": 0.19783493892144205,
+      "grad_norm": 0.6418655391021179,
+      "learning_rate": 9.280506069230945e-06,
+      "loss": 0.0139,
+      "step": 996
+    },
+    {
+      "epoch": 0.19803356837819047,
+      "grad_norm": 0.6491724713923708,
+      "learning_rate": 9.278842339119388e-06,
+      "loss": 0.0159,
+      "step": 997
+    },
+    {
+      "epoch": 0.19823219783493892,
+      "grad_norm": 0.40209317799866046,
+      "learning_rate": 9.277176837143501e-06,
+      "loss": 0.0074,
+      "step": 998
+    },
+    {
+      "epoch": 0.19843082729168735,
+      "grad_norm": 0.5920855302131218,
+      "learning_rate": 9.27550956399297e-06,
+      "loss": 0.023,
+      "step": 999
+    },
+    {
+      "epoch": 0.1986294567484358,
+      "grad_norm": 0.37861800255788186,
+      "learning_rate": 9.27384052035821e-06,
+      "loss": 0.0131,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19882808620518422,
+      "grad_norm": 0.4811422211153491,
+      "learning_rate": 9.27216970693037e-06,
+      "loss": 0.0152,
+      "step": 1001
+    },
+    {
+      "epoch": 0.19902671566193267,
+      "grad_norm": 0.6166117246588404,
+      "learning_rate": 9.270497124401332e-06,
+      "loss": 0.0159,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1992253451186811,
+      "grad_norm": 0.6776544917793748,
+      "learning_rate": 9.268822773463715e-06,
+      "loss": 0.014,
+      "step": 1003
+    },
+    {
+      "epoch": 0.19942397457542954,
+      "grad_norm": 0.3906605633332408,
+      "learning_rate": 9.267146654810859e-06,
+      "loss": 0.0134,
+      "step": 1004
+    },
+    {
+      "epoch": 0.19962260403217796,
+      "grad_norm": 0.3221183282074231,
+      "learning_rate": 9.265468769136847e-06,
+      "loss": 0.0117,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19982123348892641,
+      "grad_norm": 0.3942134464598503,
+      "learning_rate": 9.26378911713649e-06,
+      "loss": 0.0114,
+      "step": 1006
+    },
+    {
+      "epoch": 0.20001986294567484,
+      "grad_norm": 0.875008215072221,
+      "learning_rate": 9.262107699505329e-06,
+      "loss": 0.0226,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2002184924024233,
+      "grad_norm": 0.3797025965662609,
+      "learning_rate": 9.260424516939636e-06,
+      "loss": 0.0122,
+      "step": 1008
+    },
+    {
+      "epoch": 0.2004171218591717,
+      "grad_norm": 0.7343719704793019,
+      "learning_rate": 9.25873957013642e-06,
+      "loss": 0.0145,
+      "step": 1009
+    },
+    {
+      "epoch": 0.20061575131592016,
+      "grad_norm": 0.536506479455833,
+      "learning_rate": 9.257052859793412e-06,
+      "loss": 0.0143,
+      "step": 1010
+    },
+    {
+      "epoch": 0.20081438077266858,
+      "grad_norm": 0.5622360593336543,
+      "learning_rate": 9.255364386609077e-06,
+      "loss": 0.0147,
+      "step": 1011
+    },
+    {
+      "epoch": 0.20101301022941703,
+      "grad_norm": 0.4163313975119902,
+      "learning_rate": 9.253674151282612e-06,
+      "loss": 0.0127,
+      "step": 1012
+    },
+    {
+      "epoch": 0.20121163968616546,
+      "grad_norm": 0.5537476076401138,
+      "learning_rate": 9.25198215451394e-06,
+      "loss": 0.0142,
+      "step": 1013
+    },
+    {
+      "epoch": 0.2014102691429139,
+      "grad_norm": 0.6048080720095486,
+      "learning_rate": 9.250288397003715e-06,
+      "loss": 0.0128,
+      "step": 1014
+    },
+    {
+      "epoch": 0.20160889859966233,
+      "grad_norm": 0.9243481861891079,
+      "learning_rate": 9.248592879453323e-06,
+      "loss": 0.024,
+      "step": 1015
+    },
+    {
+      "epoch": 0.20180752805641078,
+      "grad_norm": 0.3405567732176348,
+      "learning_rate": 9.246895602564874e-06,
+      "loss": 0.0085,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2020061575131592,
+      "grad_norm": 0.8790274897891412,
+      "learning_rate": 9.245196567041207e-06,
+      "loss": 0.0128,
+      "step": 1017
+    },
+    {
+      "epoch": 0.20220478696990762,
+      "grad_norm": 0.6058890979082735,
+      "learning_rate": 9.243495773585896e-06,
+      "loss": 0.0165,
+      "step": 1018
+    },
+    {
+      "epoch": 0.20240341642665607,
+      "grad_norm": 0.3990747368828865,
+      "learning_rate": 9.241793222903233e-06,
+      "loss": 0.0131,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2026020458834045,
+      "grad_norm": 0.4972091523991518,
+      "learning_rate": 9.240088915698243e-06,
+      "loss": 0.0152,
+      "step": 1020
+    },
+    {
+      "epoch": 0.20280067534015295,
+      "grad_norm": 0.9145107437455913,
+      "learning_rate": 9.238382852676679e-06,
+      "loss": 0.013,
+      "step": 1021
+    },
+    {
+      "epoch": 0.20299930479690137,
+      "grad_norm": 0.7916389183212782,
+      "learning_rate": 9.236675034545022e-06,
+      "loss": 0.0158,
+      "step": 1022
+    },
+    {
+      "epoch": 0.20319793425364982,
+      "grad_norm": 1.312567937111962,
+      "learning_rate": 9.234965462010475e-06,
+      "loss": 0.0235,
+      "step": 1023
+    },
+    {
+      "epoch": 0.20339656371039824,
+      "grad_norm": 0.7879443556012908,
+      "learning_rate": 9.233254135780973e-06,
+      "loss": 0.0134,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2035951931671467,
+      "grad_norm": 0.7144955186930526,
+      "learning_rate": 9.23154105656517e-06,
+      "loss": 0.0185,
+      "step": 1025
+    },
+    {
+      "epoch": 0.20379382262389512,
+      "grad_norm": 0.8772618199422003,
+      "learning_rate": 9.229826225072455e-06,
+      "loss": 0.0115,
+      "step": 1026
+    },
+    {
+      "epoch": 0.20399245208064357,
+      "grad_norm": 0.3508519745307009,
+      "learning_rate": 9.228109642012934e-06,
+      "loss": 0.0161,
+      "step": 1027
+    },
+    {
+      "epoch": 0.204191081537392,
+      "grad_norm": 1.142282664161455,
+      "learning_rate": 9.226391308097446e-06,
+      "loss": 0.0219,
+      "step": 1028
+    },
+    {
+      "epoch": 0.20438971099414044,
+      "grad_norm": 0.6781354611906164,
+      "learning_rate": 9.22467122403755e-06,
+      "loss": 0.0236,
+      "step": 1029
+    },
+    {
+      "epoch": 0.20458834045088886,
+      "grad_norm": 0.7331441338639288,
+      "learning_rate": 9.22294939054553e-06,
+      "loss": 0.0144,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2047869699076373,
+      "grad_norm": 0.5178653186934357,
+      "learning_rate": 9.221225808334396e-06,
+      "loss": 0.0142,
+      "step": 1031
+    },
+    {
+      "epoch": 0.20498559936438573,
+      "grad_norm": 0.2608341992646028,
+      "learning_rate": 9.219500478117883e-06,
+      "loss": 0.0155,
+      "step": 1032
+    },
+    {
+      "epoch": 0.20518422882113418,
+      "grad_norm": 1.2106616895089297,
+      "learning_rate": 9.217773400610447e-06,
+      "loss": 0.0175,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2053828582778826,
+      "grad_norm": 0.28032336350734227,
+      "learning_rate": 9.21604457652727e-06,
+      "loss": 0.0128,
+      "step": 1034
+    },
+    {
+      "epoch": 0.20558148773463106,
+      "grad_norm": 0.6118204505736901,
+      "learning_rate": 9.214314006584256e-06,
+      "loss": 0.0181,
+      "step": 1035
+    },
+    {
+      "epoch": 0.20578011719137948,
+      "grad_norm": 0.579822360863569,
+      "learning_rate": 9.21258169149803e-06,
+      "loss": 0.0183,
+      "step": 1036
+    },
+    {
+      "epoch": 0.20597874664812793,
+      "grad_norm": 0.4047362585138213,
+      "learning_rate": 9.210847631985946e-06,
+      "loss": 0.0095,
+      "step": 1037
+    },
+    {
+      "epoch": 0.20617737610487635,
+      "grad_norm": 0.9852053076273041,
+      "learning_rate": 9.209111828766075e-06,
+      "loss": 0.0219,
+      "step": 1038
+    },
+    {
+      "epoch": 0.20637600556162478,
+      "grad_norm": 0.31297991161612315,
+      "learning_rate": 9.207374282557211e-06,
+      "loss": 0.0117,
+      "step": 1039
+    },
+    {
+      "epoch": 0.20657463501837323,
+      "grad_norm": 0.4164929500440738,
+      "learning_rate": 9.20563499407887e-06,
+      "loss": 0.0077,
+      "step": 1040
+    },
+    {
+      "epoch": 0.20677326447512165,
+      "grad_norm": 0.3180467674371223,
+      "learning_rate": 9.203893964051287e-06,
+      "loss": 0.011,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2069718939318701,
+      "grad_norm": 1.0524726224029894,
+      "learning_rate": 9.202151193195426e-06,
+      "loss": 0.0217,
+      "step": 1042
+    },
+    {
+      "epoch": 0.20717052338861852,
+      "grad_norm": 0.8350304497147369,
+      "learning_rate": 9.200406682232962e-06,
+      "loss": 0.0206,
+      "step": 1043
+    },
+    {
+      "epoch": 0.20736915284536697,
+      "grad_norm": 0.5099630567366636,
+      "learning_rate": 9.198660431886299e-06,
+      "loss": 0.0114,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2075677823021154,
+      "grad_norm": 0.45140417770852326,
+      "learning_rate": 9.196912442878555e-06,
+      "loss": 0.0187,
+      "step": 1045
+    },
+    {
+      "epoch": 0.20776641175886384,
+      "grad_norm": 0.3877058148364568,
+      "learning_rate": 9.195162715933573e-06,
+      "loss": 0.0146,
+      "step": 1046
+    },
+    {
+      "epoch": 0.20796504121561227,
+      "grad_norm": 0.5643704108621737,
+      "learning_rate": 9.19341125177591e-06,
+      "loss": 0.0204,
+      "step": 1047
+    },
+    {
+      "epoch": 0.20816367067236072,
+      "grad_norm": 1.1260233639536537,
+      "learning_rate": 9.191658051130845e-06,
+      "loss": 0.0176,
+      "step": 1048
+    },
+    {
+      "epoch": 0.20836230012910914,
+      "grad_norm": 0.8511847732487132,
+      "learning_rate": 9.189903114724382e-06,
+      "loss": 0.0185,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2085609295858576,
+      "grad_norm": 1.1694023155755944,
+      "learning_rate": 9.188146443283233e-06,
+      "loss": 0.02,
+      "step": 1050
+    },
+    {
+      "epoch": 0.208759559042606,
+      "grad_norm": 0.5304190834875754,
+      "learning_rate": 9.186388037534836e-06,
+      "loss": 0.0124,
+      "step": 1051
+    },
+    {
+      "epoch": 0.20895818849935446,
+      "grad_norm": 0.9189842414829346,
+      "learning_rate": 9.184627898207346e-06,
+      "loss": 0.0145,
+      "step": 1052
+    },
+    {
+      "epoch": 0.2091568179561029,
+      "grad_norm": 0.6809255153940301,
+      "learning_rate": 9.182866026029633e-06,
+      "loss": 0.0188,
+      "step": 1053
+    },
+    {
+      "epoch": 0.20935544741285134,
+      "grad_norm": 0.36494308828456623,
+      "learning_rate": 9.181102421731289e-06,
+      "loss": 0.0146,
+      "step": 1054
+    },
+    {
+      "epoch": 0.20955407686959976,
+      "grad_norm": 0.734816285805914,
+      "learning_rate": 9.179337086042618e-06,
+      "loss": 0.018,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2097527063263482,
+      "grad_norm": 0.734510701009242,
+      "learning_rate": 9.177570019694646e-06,
+      "loss": 0.0149,
+      "step": 1056
+    },
+    {
+      "epoch": 0.20995133578309663,
+      "grad_norm": 0.3903628569085709,
+      "learning_rate": 9.175801223419111e-06,
+      "loss": 0.0095,
+      "step": 1057
+    },
+    {
+      "epoch": 0.21014996523984508,
+      "grad_norm": 0.18988569699286045,
+      "learning_rate": 9.174030697948472e-06,
+      "loss": 0.0053,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2103485946965935,
+      "grad_norm": 0.22752236325582623,
+      "learning_rate": 9.1722584440159e-06,
+      "loss": 0.0103,
+      "step": 1059
+    },
+    {
+      "epoch": 0.21054722415334193,
+      "grad_norm": 0.842296179765435,
+      "learning_rate": 9.170484462355287e-06,
+      "loss": 0.0154,
+      "step": 1060
+    },
+    {
+      "epoch": 0.21074585361009038,
+      "grad_norm": 0.4916092773194938,
+      "learning_rate": 9.168708753701232e-06,
+      "loss": 0.0156,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2109444830668388,
+      "grad_norm": 0.619936716278479,
+      "learning_rate": 9.166931318789058e-06,
+      "loss": 0.0122,
+      "step": 1062
+    },
+    {
+      "epoch": 0.21114311252358725,
+      "grad_norm": 0.7419851034682392,
+      "learning_rate": 9.165152158354797e-06,
+      "loss": 0.0168,
+      "step": 1063
+    },
+    {
+      "epoch": 0.21134174198033567,
+      "grad_norm": 0.5313205892992046,
+      "learning_rate": 9.163371273135198e-06,
+      "loss": 0.016,
+      "step": 1064
+    },
+    {
+      "epoch": 0.21154037143708412,
+      "grad_norm": 1.029560156219283,
+      "learning_rate": 9.161588663867725e-06,
+      "loss": 0.0215,
+      "step": 1065
+    },
+    {
+      "epoch": 0.21173900089383255,
+      "grad_norm": 0.9348963919399064,
+      "learning_rate": 9.159804331290553e-06,
+      "loss": 0.0188,
+      "step": 1066
+    },
+    {
+      "epoch": 0.211937630350581,
+      "grad_norm": 0.3743440188348788,
+      "learning_rate": 9.158018276142573e-06,
+      "loss": 0.0098,
+      "step": 1067
+    },
+    {
+      "epoch": 0.21213625980732942,
+      "grad_norm": 1.169169214182386,
+      "learning_rate": 9.15623049916339e-06,
+      "loss": 0.0166,
+      "step": 1068
+    },
+    {
+      "epoch": 0.21233488926407787,
+      "grad_norm": 1.946374912712508,
+      "learning_rate": 9.15444100109332e-06,
+      "loss": 0.0123,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2125335187208263,
+      "grad_norm": 0.3756066632440553,
+      "learning_rate": 9.15264978267339e-06,
+      "loss": 0.017,
+      "step": 1070
+    },
+    {
+      "epoch": 0.21273214817757474,
+      "grad_norm": 0.615631571698432,
+      "learning_rate": 9.150856844645345e-06,
+      "loss": 0.0148,
+      "step": 1071
+    },
+    {
+      "epoch": 0.21293077763432317,
+      "grad_norm": 0.5711931753500064,
+      "learning_rate": 9.149062187751635e-06,
+      "loss": 0.0135,
+      "step": 1072
+    },
+    {
+      "epoch": 0.21312940709107162,
+      "grad_norm": 0.3195420308665485,
+      "learning_rate": 9.14726581273543e-06,
+      "loss": 0.0106,
+      "step": 1073
+    },
+    {
+      "epoch": 0.21332803654782004,
+      "grad_norm": 0.7675454724598806,
+      "learning_rate": 9.145467720340607e-06,
+      "loss": 0.0186,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2135266660045685,
+      "grad_norm": 0.7176331742224797,
+      "learning_rate": 9.143667911311748e-06,
+      "loss": 0.0155,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2137252954613169,
+      "grad_norm": 0.3307945238844646,
+      "learning_rate": 9.14186638639416e-06,
+      "loss": 0.0162,
+      "step": 1076
+    },
+    {
+      "epoch": 0.21392392491806536,
+      "grad_norm": 0.8244909859270414,
+      "learning_rate": 9.140063146333849e-06,
+      "loss": 0.0124,
+      "step": 1077
+    },
+    {
+      "epoch": 0.21412255437481378,
+      "grad_norm": 0.8379862052945656,
+      "learning_rate": 9.138258191877534e-06,
+      "loss": 0.0182,
+      "step": 1078
+    },
+    {
+      "epoch": 0.21432118383156223,
+      "grad_norm": 0.31645715871594615,
+      "learning_rate": 9.136451523772644e-06,
+      "loss": 0.0155,
+      "step": 1079
+    },
+    {
+      "epoch": 0.21451981328831066,
+      "grad_norm": 0.573196275828462,
+      "learning_rate": 9.134643142767324e-06,
+      "loss": 0.011,
+      "step": 1080
+    },
+    {
+      "epoch": 0.21471844274505908,
+      "grad_norm": 0.7554029631929728,
+      "learning_rate": 9.132833049610417e-06,
+      "loss": 0.0152,
+      "step": 1081
+    },
+    {
+      "epoch": 0.21491707220180753,
+      "grad_norm": 0.45963932693629606,
+      "learning_rate": 9.131021245051482e-06,
+      "loss": 0.0086,
+      "step": 1082
+    },
+    {
+      "epoch": 0.21511570165855595,
+      "grad_norm": 0.3667521583026654,
+      "learning_rate": 9.129207729840787e-06,
+      "loss": 0.0151,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2153143311153044,
+      "grad_norm": 0.7758394373843591,
+      "learning_rate": 9.127392504729308e-06,
+      "loss": 0.0119,
+      "step": 1084
+    },
+    {
+      "epoch": 0.21551296057205283,
+      "grad_norm": 0.8577031176131009,
+      "learning_rate": 9.125575570468726e-06,
+      "loss": 0.0189,
+      "step": 1085
+    },
+    {
+      "epoch": 0.21571159002880128,
+      "grad_norm": 0.4101225829842196,
+      "learning_rate": 9.123756927811429e-06,
+      "loss": 0.0147,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2159102194855497,
+      "grad_norm": 0.20450257055095583,
+      "learning_rate": 9.12193657751052e-06,
+      "loss": 0.0097,
+      "step": 1087
+    },
+    {
+      "epoch": 0.21610884894229815,
+      "grad_norm": 0.7261120497720621,
+      "learning_rate": 9.120114520319801e-06,
+      "loss": 0.0134,
+      "step": 1088
+    },
+    {
+      "epoch": 0.21630747839904657,
+      "grad_norm": 1.0613311521030548,
+      "learning_rate": 9.118290756993787e-06,
+      "loss": 0.0204,
+      "step": 1089
+    },
+    {
+      "epoch": 0.21650610785579502,
+      "grad_norm": 0.852239674073818,
+      "learning_rate": 9.116465288287693e-06,
+      "loss": 0.0327,
+      "step": 1090
+    },
+    {
+      "epoch": 0.21670473731254344,
+      "grad_norm": 0.682837045981,
+      "learning_rate": 9.114638114957444e-06,
+      "loss": 0.0151,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2169033667692919,
+      "grad_norm": 0.5746264192915811,
+      "learning_rate": 9.112809237759675e-06,
+      "loss": 0.0167,
+      "step": 1092
+    },
+    {
+      "epoch": 0.21710199622604032,
+      "grad_norm": 0.5489395595406013,
+      "learning_rate": 9.110978657451716e-06,
+      "loss": 0.0104,
+      "step": 1093
+    },
+    {
+      "epoch": 0.21730062568278877,
+      "grad_norm": 1.0914265746523228,
+      "learning_rate": 9.109146374791615e-06,
+      "loss": 0.0195,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2174992551395372,
+      "grad_norm": 1.041328850309126,
+      "learning_rate": 9.107312390538114e-06,
+      "loss": 0.0177,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21769788459628564,
+      "grad_norm": 1.047327530765551,
+      "learning_rate": 9.105476705450666e-06,
+      "loss": 0.0151,
+      "step": 1096
+    },
+    {
+      "epoch": 0.21789651405303406,
+      "grad_norm": 0.7382313477764061,
+      "learning_rate": 9.103639320289424e-06,
+      "loss": 0.0116,
+      "step": 1097
+    },
+    {
+      "epoch": 0.2180951435097825,
+      "grad_norm": 0.28713859066703235,
+      "learning_rate": 9.10180023581525e-06,
+      "loss": 0.0136,
+      "step": 1098
+    },
+    {
+      "epoch": 0.21829377296653094,
+      "grad_norm": 1.132089412775031,
+      "learning_rate": 9.099959452789706e-06,
+      "loss": 0.0177,
+      "step": 1099
+    },
+    {
+      "epoch": 0.21849240242327939,
+      "grad_norm": 0.708856856834622,
+      "learning_rate": 9.098116971975058e-06,
+      "loss": 0.0123,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2186910318800278,
+      "grad_norm": 0.6004021170139507,
+      "learning_rate": 9.096272794134276e-06,
+      "loss": 0.0214,
+      "step": 1101
+    },
+    {
+      "epoch": 0.21888966133677623,
+      "grad_norm": 1.5930865385924153,
+      "learning_rate": 9.094426920031033e-06,
+      "loss": 0.0292,
+      "step": 1102
+    },
+    {
+      "epoch": 0.21908829079352468,
+      "grad_norm": 0.7947683989906318,
+      "learning_rate": 9.092579350429703e-06,
+      "loss": 0.014,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2192869202502731,
+      "grad_norm": 0.9955767036580443,
+      "learning_rate": 9.090730086095359e-06,
+      "loss": 0.0161,
+      "step": 1104
+    },
+    {
+      "epoch": 0.21948554970702155,
+      "grad_norm": 0.3082258407552644,
+      "learning_rate": 9.088879127793786e-06,
+      "loss": 0.0102,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21968417916376998,
+      "grad_norm": 0.7479627577909793,
+      "learning_rate": 9.08702647629146e-06,
+      "loss": 0.01,
+      "step": 1106
+    },
+    {
+      "epoch": 0.21988280862051843,
+      "grad_norm": 1.2508502872987448,
+      "learning_rate": 9.085172132355563e-06,
+      "loss": 0.0181,
+      "step": 1107
+    },
+    {
+      "epoch": 0.22008143807726685,
+      "grad_norm": 1.3082638036499745,
+      "learning_rate": 9.083316096753979e-06,
+      "loss": 0.0182,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2202800675340153,
+      "grad_norm": 0.48875456172981985,
+      "learning_rate": 9.081458370255285e-06,
+      "loss": 0.0155,
+      "step": 1109
+    },
+    {
+      "epoch": 0.22047869699076372,
+      "grad_norm": 0.43843962541786946,
+      "learning_rate": 9.079598953628769e-06,
+      "loss": 0.0118,
+      "step": 1110
+    },
+    {
+      "epoch": 0.22067732644751217,
+      "grad_norm": 0.692522124440947,
+      "learning_rate": 9.077737847644411e-06,
+      "loss": 0.0168,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2208759559042606,
+      "grad_norm": 0.5909955076987183,
+      "learning_rate": 9.075875053072895e-06,
+      "loss": 0.0146,
+      "step": 1112
+    },
+    {
+      "epoch": 0.22107458536100905,
+      "grad_norm": 0.8872938011038043,
+      "learning_rate": 9.0740105706856e-06,
+      "loss": 0.0168,
+      "step": 1113
+    },
+    {
+      "epoch": 0.22127321481775747,
+      "grad_norm": 0.44692462034582914,
+      "learning_rate": 9.072144401254607e-06,
+      "loss": 0.0096,
+      "step": 1114
+    },
+    {
+      "epoch": 0.22147184427450592,
+      "grad_norm": 0.8271646623948226,
+      "learning_rate": 9.070276545552696e-06,
+      "loss": 0.0113,
+      "step": 1115
+    },
+    {
+      "epoch": 0.22167047373125434,
+      "grad_norm": 0.46981013701744095,
+      "learning_rate": 9.068407004353346e-06,
+      "loss": 0.0091,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2218691031880028,
+      "grad_norm": 0.6096189754072621,
+      "learning_rate": 9.066535778430727e-06,
+      "loss": 0.0106,
+      "step": 1117
+    },
+    {
+      "epoch": 0.22206773264475121,
+      "grad_norm": 1.2355353818301986,
+      "learning_rate": 9.064662868559714e-06,
+      "loss": 0.0177,
+      "step": 1118
+    },
+    {
+      "epoch": 0.22226636210149966,
+      "grad_norm": 0.7921361207312948,
+      "learning_rate": 9.062788275515878e-06,
+      "loss": 0.0117,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2224649915582481,
+      "grad_norm": 0.621118395225743,
+      "learning_rate": 9.060912000075489e-06,
+      "loss": 0.0184,
+      "step": 1120
+    },
+    {
+      "epoch": 0.22266362101499654,
+      "grad_norm": 1.2199946768983834,
+      "learning_rate": 9.059034043015505e-06,
+      "loss": 0.022,
+      "step": 1121
+    },
+    {
+      "epoch": 0.22286225047174496,
+      "grad_norm": 1.200785379496156,
+      "learning_rate": 9.057154405113588e-06,
+      "loss": 0.0146,
+      "step": 1122
+    },
+    {
+      "epoch": 0.22306087992849338,
+      "grad_norm": 0.7879812928114914,
+      "learning_rate": 9.055273087148095e-06,
+      "loss": 0.0154,
+      "step": 1123
+    },
+    {
+      "epoch": 0.22325950938524183,
+      "grad_norm": 0.865383817784091,
+      "learning_rate": 9.053390089898078e-06,
+      "loss": 0.0151,
+      "step": 1124
+    },
+    {
+      "epoch": 0.22345813884199026,
+      "grad_norm": 0.6021432751244368,
+      "learning_rate": 9.051505414143283e-06,
+      "loss": 0.0213,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2236567682987387,
+      "grad_norm": 0.7551185452295719,
+      "learning_rate": 9.049619060664155e-06,
+      "loss": 0.0234,
+      "step": 1126
+    },
+    {
+      "epoch": 0.22385539775548713,
+      "grad_norm": 0.9095260202243808,
+      "learning_rate": 9.047731030241827e-06,
+      "loss": 0.0119,
+      "step": 1127
+    },
+    {
+      "epoch": 0.22405402721223558,
+      "grad_norm": 0.7060999713583611,
+      "learning_rate": 9.045841323658136e-06,
+      "loss": 0.0139,
+      "step": 1128
+    },
+    {
+      "epoch": 0.224252656668984,
+      "grad_norm": 0.611769741942482,
+      "learning_rate": 9.043949941695602e-06,
+      "loss": 0.0135,
+      "step": 1129
+    },
+    {
+      "epoch": 0.22445128612573245,
+      "grad_norm": 0.6041679576728929,
+      "learning_rate": 9.042056885137447e-06,
+      "loss": 0.0208,
+      "step": 1130
+    },
+    {
+      "epoch": 0.22464991558248087,
+      "grad_norm": 0.4868808950406833,
+      "learning_rate": 9.040162154767585e-06,
+      "loss": 0.0139,
+      "step": 1131
+    },
+    {
+      "epoch": 0.22484854503922933,
+      "grad_norm": 0.37898547751832734,
+      "learning_rate": 9.03826575137062e-06,
+      "loss": 0.0134,
+      "step": 1132
+    },
+    {
+      "epoch": 0.22504717449597775,
+      "grad_norm": 0.3871203326790516,
+      "learning_rate": 9.036367675731852e-06,
+      "loss": 0.014,
+      "step": 1133
+    },
+    {
+      "epoch": 0.2252458039527262,
+      "grad_norm": 0.4158985639649831,
+      "learning_rate": 9.03446792863727e-06,
+      "loss": 0.0155,
+      "step": 1134
+    },
+    {
+      "epoch": 0.22544443340947462,
+      "grad_norm": 0.8248065463728428,
+      "learning_rate": 9.03256651087356e-06,
+      "loss": 0.0191,
+      "step": 1135
+    },
+    {
+      "epoch": 0.22564306286622307,
+      "grad_norm": 0.6997112013650323,
+      "learning_rate": 9.030663423228096e-06,
+      "loss": 0.0119,
+      "step": 1136
+    },
+    {
+      "epoch": 0.2258416923229715,
+      "grad_norm": 0.40813675543898414,
+      "learning_rate": 9.028758666488946e-06,
+      "loss": 0.0106,
+      "step": 1137
+    },
+    {
+      "epoch": 0.22604032177971994,
+      "grad_norm": 0.27264796677090325,
+      "learning_rate": 9.026852241444865e-06,
+      "loss": 0.0117,
+      "step": 1138
+    },
+    {
+      "epoch": 0.22623895123646837,
+      "grad_norm": 0.7634799874549393,
+      "learning_rate": 9.024944148885305e-06,
+      "loss": 0.0137,
+      "step": 1139
+    },
+    {
+      "epoch": 0.22643758069321682,
+      "grad_norm": 0.37311413583560454,
+      "learning_rate": 9.023034389600403e-06,
+      "loss": 0.014,
+      "step": 1140
+    },
+    {
+      "epoch": 0.22663621014996524,
+      "grad_norm": 1.1650537636675808,
+      "learning_rate": 9.021122964380988e-06,
+      "loss": 0.0149,
+      "step": 1141
+    },
+    {
+      "epoch": 0.22683483960671366,
+      "grad_norm": 0.6052730792928078,
+      "learning_rate": 9.019209874018581e-06,
+      "loss": 0.0153,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2270334690634621,
+      "grad_norm": 0.6627925620828036,
+      "learning_rate": 9.01729511930539e-06,
+      "loss": 0.0132,
+      "step": 1143
+    },
+    {
+      "epoch": 0.22723209852021053,
+      "grad_norm": 1.831246147791384,
+      "learning_rate": 9.015378701034315e-06,
+      "loss": 0.0222,
+      "step": 1144
+    },
+    {
+      "epoch": 0.22743072797695899,
+      "grad_norm": 0.6830687198229002,
+      "learning_rate": 9.013460619998937e-06,
+      "loss": 0.0095,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2276293574337074,
+      "grad_norm": 0.6897632451876908,
+      "learning_rate": 9.011540876993539e-06,
+      "loss": 0.0145,
+      "step": 1146
+    },
+    {
+      "epoch": 0.22782798689045586,
+      "grad_norm": 0.45414406716937405,
+      "learning_rate": 9.00961947281308e-06,
+      "loss": 0.0135,
+      "step": 1147
+    },
+    {
+      "epoch": 0.22802661634720428,
+      "grad_norm": 0.6724858979266666,
+      "learning_rate": 9.007696408253212e-06,
+      "loss": 0.0176,
+      "step": 1148
+    },
+    {
+      "epoch": 0.22822524580395273,
+      "grad_norm": 0.5585341592638718,
+      "learning_rate": 9.005771684110275e-06,
+      "loss": 0.0142,
+      "step": 1149
+    },
+    {
+      "epoch": 0.22842387526070115,
+      "grad_norm": 0.33631448656179974,
+      "learning_rate": 9.003845301181296e-06,
+      "loss": 0.0104,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2286225047174496,
+      "grad_norm": 0.936608731900455,
+      "learning_rate": 9.001917260263986e-06,
+      "loss": 0.0176,
+      "step": 1151
+    },
+    {
+      "epoch": 0.22882113417419803,
+      "grad_norm": 0.450082775257085,
+      "learning_rate": 8.999987562156747e-06,
+      "loss": 0.0108,
+      "step": 1152
+    },
+    {
+      "epoch": 0.22901976363094648,
+      "grad_norm": 0.6502188418132533,
+      "learning_rate": 8.998056207658662e-06,
+      "loss": 0.0197,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2292183930876949,
+      "grad_norm": 0.7819061233264434,
+      "learning_rate": 8.996123197569508e-06,
+      "loss": 0.0116,
+      "step": 1154
+    },
+    {
+      "epoch": 0.22941702254444335,
+      "grad_norm": 0.8372844196296657,
+      "learning_rate": 8.994188532689739e-06,
+      "loss": 0.0124,
+      "step": 1155
+    },
+    {
+      "epoch": 0.22961565200119177,
+      "grad_norm": 0.4402518430330382,
+      "learning_rate": 8.992252213820498e-06,
+      "loss": 0.0117,
+      "step": 1156
+    },
+    {
+      "epoch": 0.22981428145794022,
+      "grad_norm": 0.3109619518398801,
+      "learning_rate": 8.990314241763614e-06,
+      "loss": 0.0105,
+      "step": 1157
+    },
+    {
+      "epoch": 0.23001291091468865,
+      "grad_norm": 0.940067780887533,
+      "learning_rate": 8.988374617321597e-06,
+      "loss": 0.0186,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2302115403714371,
+      "grad_norm": 1.0503338550839172,
+      "learning_rate": 8.986433341297646e-06,
+      "loss": 0.0221,
+      "step": 1159
+    },
+    {
+      "epoch": 0.23041016982818552,
+      "grad_norm": 0.6966779242628477,
+      "learning_rate": 8.984490414495642e-06,
+      "loss": 0.0104,
+      "step": 1160
+    },
+    {
+      "epoch": 0.23060879928493397,
+      "grad_norm": 0.29127162723599453,
+      "learning_rate": 8.982545837720148e-06,
+      "loss": 0.008,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2308074287416824,
+      "grad_norm": 0.7864175276944835,
+      "learning_rate": 8.980599611776408e-06,
+      "loss": 0.0203,
+      "step": 1162
+    },
+    {
+      "epoch": 0.2310060581984308,
+      "grad_norm": 0.6641316865126137,
+      "learning_rate": 8.978651737470354e-06,
+      "loss": 0.016,
+      "step": 1163
+    },
+    {
+      "epoch": 0.23120468765517926,
+      "grad_norm": 0.9355779041765688,
+      "learning_rate": 8.976702215608603e-06,
+      "loss": 0.0196,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2314033171119277,
+      "grad_norm": 0.5906387589001846,
+      "learning_rate": 8.974751046998445e-06,
+      "loss": 0.0188,
+      "step": 1165
+    },
+    {
+      "epoch": 0.23160194656867614,
+      "grad_norm": 1.0575477626994387,
+      "learning_rate": 8.97279823244786e-06,
+      "loss": 0.0118,
+      "step": 1166
+    },
+    {
+      "epoch": 0.23180057602542456,
+      "grad_norm": 0.2202881228331673,
+      "learning_rate": 8.970843772765505e-06,
+      "loss": 0.0091,
+      "step": 1167
+    },
+    {
+      "epoch": 0.231999205482173,
+      "grad_norm": 0.7075656613422315,
+      "learning_rate": 8.968887668760719e-06,
+      "loss": 0.0158,
+      "step": 1168
+    },
+    {
+      "epoch": 0.23219783493892143,
+      "grad_norm": 0.2683433024299603,
+      "learning_rate": 8.966929921243526e-06,
+      "loss": 0.012,
+      "step": 1169
+    },
+    {
+      "epoch": 0.23239646439566988,
+      "grad_norm": 0.42237177538366844,
+      "learning_rate": 8.964970531024624e-06,
+      "loss": 0.0131,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2325950938524183,
+      "grad_norm": 0.594094475622576,
+      "learning_rate": 8.963009498915396e-06,
+      "loss": 0.0119,
+      "step": 1171
+    },
+    {
+      "epoch": 0.23279372330916676,
+      "grad_norm": 0.406385112438069,
+      "learning_rate": 8.961046825727904e-06,
+      "loss": 0.0113,
+      "step": 1172
+    },
+    {
+      "epoch": 0.23299235276591518,
+      "grad_norm": 0.8033200609983158,
+      "learning_rate": 8.959082512274885e-06,
+      "loss": 0.0123,
+      "step": 1173
+    },
+    {
+      "epoch": 0.23319098222266363,
+      "grad_norm": 0.47018932211309783,
+      "learning_rate": 8.957116559369767e-06,
+      "loss": 0.0144,
+      "step": 1174
+    },
+    {
+      "epoch": 0.23338961167941205,
+      "grad_norm": 0.6077335509097652,
+      "learning_rate": 8.955148967826642e-06,
+      "loss": 0.0191,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2335882411361605,
+      "grad_norm": 0.8460343255845043,
+      "learning_rate": 8.95317973846029e-06,
+      "loss": 0.0231,
+      "step": 1176
+    },
+    {
+      "epoch": 0.23378687059290892,
+      "grad_norm": 0.568144005560869,
+      "learning_rate": 8.951208872086166e-06,
+      "loss": 0.0148,
+      "step": 1177
+    },
+    {
+      "epoch": 0.23398550004965737,
+      "grad_norm": 0.6107991897583083,
+      "learning_rate": 8.949236369520406e-06,
+      "loss": 0.0184,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2341841295064058,
+      "grad_norm": 0.42161119081345094,
+      "learning_rate": 8.947262231579822e-06,
+      "loss": 0.0119,
+      "step": 1179
+    },
+    {
+      "epoch": 0.23438275896315425,
+      "grad_norm": 0.40846557188713406,
+      "learning_rate": 8.945286459081899e-06,
+      "loss": 0.0137,
+      "step": 1180
+    },
+    {
+      "epoch": 0.23458138841990267,
+      "grad_norm": 0.763309241318152,
+      "learning_rate": 8.943309052844806e-06,
+      "loss": 0.0134,
+      "step": 1181
+    },
+    {
+      "epoch": 0.23478001787665112,
+      "grad_norm": 0.5803819518194214,
+      "learning_rate": 8.941330013687382e-06,
+      "loss": 0.0179,
+      "step": 1182
+    },
+    {
+      "epoch": 0.23497864733339954,
+      "grad_norm": 1.039753700867902,
+      "learning_rate": 8.939349342429144e-06,
+      "loss": 0.0152,
+      "step": 1183
+    },
+    {
+      "epoch": 0.23517727679014797,
+      "grad_norm": 0.5593224702420636,
+      "learning_rate": 8.937367039890291e-06,
+      "loss": 0.0138,
+      "step": 1184
+    },
+    {
+      "epoch": 0.23537590624689642,
+      "grad_norm": 0.6151456252694801,
+      "learning_rate": 8.93538310689169e-06,
+      "loss": 0.0202,
+      "step": 1185
+    },
+    {
+      "epoch": 0.23557453570364484,
+      "grad_norm": 0.2979372338609462,
+      "learning_rate": 8.933397544254884e-06,
+      "loss": 0.0121,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2357731651603933,
+      "grad_norm": 1.520553652264108,
+      "learning_rate": 8.931410352802095e-06,
+      "loss": 0.0222,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2359717946171417,
+      "grad_norm": 0.504693883508692,
+      "learning_rate": 8.929421533356215e-06,
+      "loss": 0.0155,
+      "step": 1188
+    },
+    {
+      "epoch": 0.23617042407389016,
+      "grad_norm": 0.7873301547010652,
+      "learning_rate": 8.927431086740814e-06,
+      "loss": 0.0253,
+      "step": 1189
+    },
+    {
+      "epoch": 0.23636905353063858,
+      "grad_norm": 0.3650087454226798,
+      "learning_rate": 8.925439013780131e-06,
+      "loss": 0.0129,
+      "step": 1190
+    },
+    {
+      "epoch": 0.23656768298738703,
+      "grad_norm": 0.8064537953076155,
+      "learning_rate": 8.923445315299085e-06,
+      "loss": 0.0157,
+      "step": 1191
+    },
+    {
+      "epoch": 0.23676631244413546,
+      "grad_norm": 0.37122234928734354,
+      "learning_rate": 8.921449992123264e-06,
+      "loss": 0.0116,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2369649419008839,
+      "grad_norm": 1.2902752115216132,
+      "learning_rate": 8.919453045078927e-06,
+      "loss": 0.0227,
+      "step": 1193
+    },
+    {
+      "epoch": 0.23716357135763233,
+      "grad_norm": 1.7179509273444584,
+      "learning_rate": 8.917454474993008e-06,
+      "loss": 0.0245,
+      "step": 1194
+    },
+    {
+      "epoch": 0.23736220081438078,
+      "grad_norm": 0.38492446902743044,
+      "learning_rate": 8.915454282693116e-06,
+      "loss": 0.0145,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2375608302711292,
+      "grad_norm": 0.4447455285152336,
+      "learning_rate": 8.913452469007526e-06,
+      "loss": 0.0125,
+      "step": 1196
+    },
+    {
+      "epoch": 0.23775945972787765,
+      "grad_norm": 0.5145154810241487,
+      "learning_rate": 8.911449034765186e-06,
+      "loss": 0.0117,
+      "step": 1197
+    },
+    {
+      "epoch": 0.23795808918462608,
+      "grad_norm": 0.3081471801818194,
+      "learning_rate": 8.90944398079572e-06,
+      "loss": 0.0184,
+      "step": 1198
+    },
+    {
+      "epoch": 0.23815671864137453,
+      "grad_norm": 0.4384280059834026,
+      "learning_rate": 8.907437307929416e-06,
+      "loss": 0.0098,
+      "step": 1199
+    },
+    {
+      "epoch": 0.23835534809812295,
+      "grad_norm": 0.3173624393698063,
+      "learning_rate": 8.905429016997236e-06,
+      "loss": 0.0154,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2385539775548714,
+      "grad_norm": 0.4241866381288965,
+      "learning_rate": 8.903419108830808e-06,
+      "loss": 0.0121,
+      "step": 1201
+    },
+    {
+      "epoch": 0.23875260701161982,
+      "grad_norm": 1.6418273480323602,
+      "learning_rate": 8.901407584262441e-06,
+      "loss": 0.0217,
+      "step": 1202
+    },
+    {
+      "epoch": 0.23895123646836827,
+      "grad_norm": 0.6482978467087832,
+      "learning_rate": 8.899394444125097e-06,
+      "loss": 0.0119,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2391498659251167,
+      "grad_norm": 0.49425790948254467,
+      "learning_rate": 8.897379689252418e-06,
+      "loss": 0.0105,
+      "step": 1204
+    },
+    {
+      "epoch": 0.23934849538186512,
+      "grad_norm": 0.7361793243822061,
+      "learning_rate": 8.895363320478715e-06,
+      "loss": 0.0167,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23954712483861357,
+      "grad_norm": 0.5247879216412412,
+      "learning_rate": 8.893345338638961e-06,
+      "loss": 0.0128,
+      "step": 1206
+    },
+    {
+      "epoch": 0.239745754295362,
+      "grad_norm": 0.7004974169321737,
+      "learning_rate": 8.891325744568802e-06,
+      "loss": 0.0124,
+      "step": 1207
+    },
+    {
+      "epoch": 0.23994438375211044,
+      "grad_norm": 0.5994164703367848,
+      "learning_rate": 8.889304539104549e-06,
+      "loss": 0.0082,
+      "step": 1208
+    },
+    {
+      "epoch": 0.24014301320885886,
+      "grad_norm": 0.6187312745317394,
+      "learning_rate": 8.887281723083179e-06,
+      "loss": 0.0143,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2403416426656073,
+      "grad_norm": 0.3401870907046762,
+      "learning_rate": 8.885257297342343e-06,
+      "loss": 0.0092,
+      "step": 1210
+    },
+    {
+      "epoch": 0.24054027212235574,
+      "grad_norm": 0.702714768582494,
+      "learning_rate": 8.883231262720348e-06,
+      "loss": 0.0149,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2407389015791042,
+      "grad_norm": 0.31545699277173606,
+      "learning_rate": 8.881203620056178e-06,
+      "loss": 0.0071,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2409375310358526,
+      "grad_norm": 0.9671126025588995,
+      "learning_rate": 8.879174370189475e-06,
+      "loss": 0.0139,
+      "step": 1213
+    },
+    {
+      "epoch": 0.24113616049260106,
+      "grad_norm": 0.9187540597662088,
+      "learning_rate": 8.87714351396055e-06,
+      "loss": 0.0264,
+      "step": 1214
+    },
+    {
+      "epoch": 0.24133478994934948,
+      "grad_norm": 0.9928211950549544,
+      "learning_rate": 8.875111052210378e-06,
+      "loss": 0.0151,
+      "step": 1215
+    },
+    {
+      "epoch": 0.24153341940609793,
+      "grad_norm": 2.2372121346405853,
+      "learning_rate": 8.873076985780602e-06,
+      "loss": 0.0327,
+      "step": 1216
+    },
+    {
+      "epoch": 0.24173204886284635,
+      "grad_norm": 0.6013185780237169,
+      "learning_rate": 8.871041315513523e-06,
+      "loss": 0.0203,
+      "step": 1217
+    },
+    {
+      "epoch": 0.2419306783195948,
+      "grad_norm": 0.4762488349863747,
+      "learning_rate": 8.869004042252111e-06,
+      "loss": 0.0185,
+      "step": 1218
+    },
+    {
+      "epoch": 0.24212930777634323,
+      "grad_norm": 0.5238008487757791,
+      "learning_rate": 8.866965166840003e-06,
+      "loss": 0.0121,
+      "step": 1219
+    },
+    {
+      "epoch": 0.24232793723309168,
+      "grad_norm": 0.8737365406833077,
+      "learning_rate": 8.864924690121489e-06,
+      "loss": 0.0154,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2425265666898401,
+      "grad_norm": 0.8165887277343837,
+      "learning_rate": 8.862882612941532e-06,
+      "loss": 0.0136,
+      "step": 1221
+    },
+    {
+      "epoch": 0.24272519614658855,
+      "grad_norm": 0.6144569559576514,
+      "learning_rate": 8.860838936145754e-06,
+      "loss": 0.0134,
+      "step": 1222
+    },
+    {
+      "epoch": 0.24292382560333697,
+      "grad_norm": 0.7154975367837693,
+      "learning_rate": 8.858793660580438e-06,
+      "loss": 0.0164,
+      "step": 1223
+    },
+    {
+      "epoch": 0.24312245506008542,
+      "grad_norm": 0.5286069304960067,
+      "learning_rate": 8.856746787092532e-06,
+      "loss": 0.0172,
+      "step": 1224
+    },
+    {
+      "epoch": 0.24332108451683385,
+      "grad_norm": 0.6339057196599259,
+      "learning_rate": 8.854698316529642e-06,
+      "loss": 0.0202,
+      "step": 1225
+    },
+    {
+      "epoch": 0.24351971397358227,
+      "grad_norm": 0.671220701801568,
+      "learning_rate": 8.852648249740041e-06,
+      "loss": 0.0151,
+      "step": 1226
+    },
+    {
+      "epoch": 0.24371834343033072,
+      "grad_norm": 1.1322941892865441,
+      "learning_rate": 8.850596587572658e-06,
+      "loss": 0.0153,
+      "step": 1227
+    },
+    {
+      "epoch": 0.24391697288707914,
+      "grad_norm": 0.5378549330329825,
+      "learning_rate": 8.848543330877084e-06,
+      "loss": 0.0186,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2441156023438276,
+      "grad_norm": 0.5673634305039723,
+      "learning_rate": 8.84648848050357e-06,
+      "loss": 0.0197,
+      "step": 1229
+    },
+    {
+      "epoch": 0.24431423180057601,
+      "grad_norm": 0.3534387414527403,
+      "learning_rate": 8.84443203730303e-06,
+      "loss": 0.0165,
+      "step": 1230
+    },
+    {
+      "epoch": 0.24451286125732447,
+      "grad_norm": 0.5672990685019907,
+      "learning_rate": 8.842374002127033e-06,
+      "loss": 0.0136,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2447114907140729,
+      "grad_norm": 0.8327923542087138,
+      "learning_rate": 8.840314375827808e-06,
+      "loss": 0.0145,
+      "step": 1232
+    },
+    {
+      "epoch": 0.24491012017082134,
+      "grad_norm": 0.3379005386095217,
+      "learning_rate": 8.838253159258245e-06,
+      "loss": 0.0112,
+      "step": 1233
+    },
+    {
+      "epoch": 0.24510874962756976,
+      "grad_norm": 0.8689521331181729,
+      "learning_rate": 8.836190353271894e-06,
+      "loss": 0.024,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2453073790843182,
+      "grad_norm": 0.42828866756204587,
+      "learning_rate": 8.834125958722958e-06,
+      "loss": 0.0115,
+      "step": 1235
+    },
+    {
+      "epoch": 0.24550600854106663,
+      "grad_norm": 0.5156919400962039,
+      "learning_rate": 8.832059976466305e-06,
+      "loss": 0.0175,
+      "step": 1236
+    },
+    {
+      "epoch": 0.24570463799781508,
+      "grad_norm": 0.6054495643078346,
+      "learning_rate": 8.82999240735745e-06,
+      "loss": 0.0098,
+      "step": 1237
+    },
+    {
+      "epoch": 0.2459032674545635,
+      "grad_norm": 0.7588058992250312,
+      "learning_rate": 8.827923252252577e-06,
+      "loss": 0.0143,
+      "step": 1238
+    },
+    {
+      "epoch": 0.24610189691131196,
+      "grad_norm": 0.38121018989369687,
+      "learning_rate": 8.825852512008518e-06,
+      "loss": 0.0122,
+      "step": 1239
+    },
+    {
+      "epoch": 0.24630052636806038,
+      "grad_norm": 0.6129760460462393,
+      "learning_rate": 8.823780187482764e-06,
+      "loss": 0.0113,
+      "step": 1240
+    },
+    {
+      "epoch": 0.24649915582480883,
+      "grad_norm": 1.2945184977220947,
+      "learning_rate": 8.821706279533465e-06,
+      "loss": 0.0135,
+      "step": 1241
+    },
+    {
+      "epoch": 0.24669778528155725,
+      "grad_norm": 0.5977689562617747,
+      "learning_rate": 8.819630789019422e-06,
+      "loss": 0.0093,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2468964147383057,
+      "grad_norm": 0.2571681909339781,
+      "learning_rate": 8.817553716800095e-06,
+      "loss": 0.0046,
+      "step": 1243
+    },
+    {
+      "epoch": 0.24709504419505413,
+      "grad_norm": 0.8248908002719997,
+      "learning_rate": 8.815475063735596e-06,
+      "loss": 0.0113,
+      "step": 1244
+    },
+    {
+      "epoch": 0.24729367365180258,
+      "grad_norm": 0.584105982723087,
+      "learning_rate": 8.813394830686695e-06,
+      "loss": 0.012,
+      "step": 1245
+    },
+    {
+      "epoch": 0.247492303108551,
+      "grad_norm": 0.6285446237309048,
+      "learning_rate": 8.811313018514812e-06,
+      "loss": 0.0189,
+      "step": 1246
+    },
+    {
+      "epoch": 0.24769093256529942,
+      "grad_norm": 0.7020367426867679,
+      "learning_rate": 8.809229628082025e-06,
+      "loss": 0.0223,
+      "step": 1247
+    },
+    {
+      "epoch": 0.24788956202204787,
+      "grad_norm": 0.7721912844272995,
+      "learning_rate": 8.807144660251065e-06,
+      "loss": 0.0201,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2480881914787963,
+      "grad_norm": 0.6697889970539281,
+      "learning_rate": 8.805058115885313e-06,
+      "loss": 0.0188,
+      "step": 1249
+    },
+    {
+      "epoch": 0.24828682093554474,
+      "grad_norm": 0.5381683696890166,
+      "learning_rate": 8.802969995848807e-06,
+      "loss": 0.0061,
+      "step": 1250
+    },
+    {
+      "epoch": 0.24848545039229317,
+      "grad_norm": 0.20343883973471363,
+      "learning_rate": 8.800880301006232e-06,
+      "loss": 0.0043,
+      "step": 1251
+    },
+    {
+      "epoch": 0.24868407984904162,
+      "grad_norm": 1.0199510305812602,
+      "learning_rate": 8.798789032222932e-06,
+      "loss": 0.0165,
+      "step": 1252
+    },
+    {
+      "epoch": 0.24888270930579004,
+      "grad_norm": 0.6275045781700873,
+      "learning_rate": 8.796696190364897e-06,
+      "loss": 0.0221,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2490813387625385,
+      "grad_norm": 1.0129544827919168,
+      "learning_rate": 8.794601776298772e-06,
+      "loss": 0.0155,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2492799682192869,
+      "grad_norm": 1.1428308322663727,
+      "learning_rate": 8.792505790891852e-06,
+      "loss": 0.0135,
+      "step": 1255
+    },
+    {
+      "epoch": 0.24947859767603536,
+      "grad_norm": 0.39150197901895456,
+      "learning_rate": 8.790408235012081e-06,
+      "loss": 0.0156,
+      "step": 1256
+    },
+    {
+      "epoch": 0.24967722713278379,
+      "grad_norm": 0.5483610810377085,
+      "learning_rate": 8.788309109528057e-06,
+      "loss": 0.0113,
+      "step": 1257
+    },
+    {
+      "epoch": 0.24987585658953224,
+      "grad_norm": 0.27854942135066457,
+      "learning_rate": 8.786208415309023e-06,
+      "loss": 0.0132,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2500744860462807,
+      "grad_norm": 0.36486249941255533,
+      "learning_rate": 8.784106153224876e-06,
+      "loss": 0.0111,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2502731155030291,
+      "grad_norm": 0.5102349401683092,
+      "learning_rate": 8.782002324146162e-06,
+      "loss": 0.0131,
+      "step": 1260
+    },
+    {
+      "epoch": 0.25047174495977753,
+      "grad_norm": 0.590643281645968,
+      "learning_rate": 8.779896928944072e-06,
+      "loss": 0.0192,
+      "step": 1261
+    },
+    {
+      "epoch": 0.25067037441652595,
+      "grad_norm": 0.6155294741908522,
+      "learning_rate": 8.777789968490449e-06,
+      "loss": 0.0108,
+      "step": 1262
+    },
+    {
+      "epoch": 0.25086900387327443,
+      "grad_norm": 0.8591597492277506,
+      "learning_rate": 8.775681443657781e-06,
+      "loss": 0.0134,
+      "step": 1263
+    },
+    {
+      "epoch": 0.25106763333002285,
+      "grad_norm": 0.36267619775257987,
+      "learning_rate": 8.773571355319213e-06,
+      "loss": 0.0091,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2512662627867713,
+      "grad_norm": 0.545487632579775,
+      "learning_rate": 8.771459704348521e-06,
+      "loss": 0.0152,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2514648922435197,
+      "grad_norm": 0.4679522780645009,
+      "learning_rate": 8.769346491620145e-06,
+      "loss": 0.0125,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2516635217002681,
+      "grad_norm": 0.5309851493131132,
+      "learning_rate": 8.767231718009161e-06,
+      "loss": 0.0105,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2518621511570166,
+      "grad_norm": 1.162362377224985,
+      "learning_rate": 8.765115384391296e-06,
+      "loss": 0.015,
+      "step": 1268
+    },
+    {
+      "epoch": 0.252060780613765,
+      "grad_norm": 0.45539747042204515,
+      "learning_rate": 8.76299749164292e-06,
+      "loss": 0.0141,
+      "step": 1269
+    },
+    {
+      "epoch": 0.25225941007051345,
+      "grad_norm": 0.5999868464347882,
+      "learning_rate": 8.76087804064105e-06,
+      "loss": 0.0171,
+      "step": 1270
+    },
+    {
+      "epoch": 0.25245803952726187,
+      "grad_norm": 0.5339189795562727,
+      "learning_rate": 8.75875703226335e-06,
+      "loss": 0.0162,
+      "step": 1271
+    },
+    {
+      "epoch": 0.25265666898401035,
+      "grad_norm": 0.859130519663411,
+      "learning_rate": 8.756634467388128e-06,
+      "loss": 0.0165,
+      "step": 1272
+    },
+    {
+      "epoch": 0.25285529844075877,
+      "grad_norm": 0.21076209185559133,
+      "learning_rate": 8.754510346894334e-06,
+      "loss": 0.0044,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2530539278975072,
+      "grad_norm": 0.7939050635268597,
+      "learning_rate": 8.752384671661566e-06,
+      "loss": 0.0182,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2532525573542556,
+      "grad_norm": 0.5766317392642453,
+      "learning_rate": 8.750257442570064e-06,
+      "loss": 0.0095,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2534511868110041,
+      "grad_norm": 0.7303525194562475,
+      "learning_rate": 8.74812866050071e-06,
+      "loss": 0.0117,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2536498162677525,
+      "grad_norm": 0.40272585071340505,
+      "learning_rate": 8.74599832633503e-06,
+      "loss": 0.0107,
+      "step": 1277
+    },
+    {
+      "epoch": 0.25384844572450094,
+      "grad_norm": 0.9286068222372685,
+      "learning_rate": 8.743866440955196e-06,
+      "loss": 0.0207,
+      "step": 1278
+    },
+    {
+      "epoch": 0.25404707518124936,
+      "grad_norm": 0.6122165638925774,
+      "learning_rate": 8.741733005244016e-06,
+      "loss": 0.0162,
+      "step": 1279
+    },
+    {
+      "epoch": 0.25424570463799784,
+      "grad_norm": 0.590624454464984,
+      "learning_rate": 8.739598020084947e-06,
+      "loss": 0.0087,
+      "step": 1280
+    },
+    {
+      "epoch": 0.25444433409474626,
+      "grad_norm": 1.016687681475864,
+      "learning_rate": 8.737461486362082e-06,
+      "loss": 0.0161,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2546429635514947,
+      "grad_norm": 0.8291330430883908,
+      "learning_rate": 8.735323404960159e-06,
+      "loss": 0.0209,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2548415930082431,
+      "grad_norm": 1.2088406629766484,
+      "learning_rate": 8.733183776764556e-06,
+      "loss": 0.016,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2550402224649916,
+      "grad_norm": 0.903068606712378,
+      "learning_rate": 8.731042602661289e-06,
+      "loss": 0.0201,
+      "step": 1284
+    },
+    {
+      "epoch": 0.25523885192174,
+      "grad_norm": 0.6420259885722851,
+      "learning_rate": 8.728899883537014e-06,
+      "loss": 0.0146,
+      "step": 1285
+    },
+    {
+      "epoch": 0.25543748137848843,
+      "grad_norm": 0.4337787015933656,
+      "learning_rate": 8.726755620279033e-06,
+      "loss": 0.0121,
+      "step": 1286
+    },
+    {
+      "epoch": 0.25563611083523685,
+      "grad_norm": 0.9125000801904051,
+      "learning_rate": 8.724609813775282e-06,
+      "loss": 0.0174,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2558347402919853,
+      "grad_norm": 0.4601365566071554,
+      "learning_rate": 8.722462464914337e-06,
+      "loss": 0.0091,
+      "step": 1288
+    },
+    {
+      "epoch": 0.25603336974873375,
+      "grad_norm": 0.6681352622131371,
+      "learning_rate": 8.720313574585412e-06,
+      "loss": 0.0168,
+      "step": 1289
+    },
+    {
+      "epoch": 0.2562319992054822,
+      "grad_norm": 0.3959505420693962,
+      "learning_rate": 8.718163143678365e-06,
+      "loss": 0.0128,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2564306286622306,
+      "grad_norm": 0.23256713254644923,
+      "learning_rate": 8.716011173083679e-06,
+      "loss": 0.0117,
+      "step": 1291
+    },
+    {
+      "epoch": 0.256629258118979,
+      "grad_norm": 0.2541092240532925,
+      "learning_rate": 8.713857663692492e-06,
+      "loss": 0.0086,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2568278875757275,
+      "grad_norm": 0.6179349251330739,
+      "learning_rate": 8.711702616396562e-06,
+      "loss": 0.014,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2570265170324759,
+      "grad_norm": 0.3453897821620428,
+      "learning_rate": 8.709546032088296e-06,
+      "loss": 0.0128,
+      "step": 1294
+    },
+    {
+      "epoch": 0.25722514648922434,
+      "grad_norm": 0.5847778245454376,
+      "learning_rate": 8.707387911660735e-06,
+      "loss": 0.0119,
+      "step": 1295
+    },
+    {
+      "epoch": 0.25742377594597277,
+      "grad_norm": 0.2704171725983602,
+      "learning_rate": 8.705228256007549e-06,
+      "loss": 0.0115,
+      "step": 1296
+    },
+    {
+      "epoch": 0.25762240540272124,
+      "grad_norm": 0.6897068421679288,
+      "learning_rate": 8.703067066023055e-06,
+      "loss": 0.0112,
+      "step": 1297
+    },
+    {
+      "epoch": 0.25782103485946967,
+      "grad_norm": 0.40321138772863513,
+      "learning_rate": 8.700904342602197e-06,
+      "loss": 0.0056,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2580196643162181,
+      "grad_norm": 0.7069935344664505,
+      "learning_rate": 8.698740086640559e-06,
+      "loss": 0.0194,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2582182937729665,
+      "grad_norm": 0.9773159570445462,
+      "learning_rate": 8.696574299034351e-06,
+      "loss": 0.0236,
+      "step": 1300
+    },
+    {
+      "epoch": 0.258416923229715,
+      "grad_norm": 1.029306053307389,
+      "learning_rate": 8.69440698068043e-06,
+      "loss": 0.0202,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2586155526864634,
+      "grad_norm": 0.59016206926503,
+      "learning_rate": 8.692238132476278e-06,
+      "loss": 0.0099,
+      "step": 1302
+    },
+    {
+      "epoch": 0.25881418214321184,
+      "grad_norm": 0.516731967140984,
+      "learning_rate": 8.690067755320012e-06,
+      "loss": 0.0094,
+      "step": 1303
+    },
+    {
+      "epoch": 0.25901281159996026,
+      "grad_norm": 0.3353964879934747,
+      "learning_rate": 8.687895850110386e-06,
+      "loss": 0.0167,
+      "step": 1304
+    },
+    {
+      "epoch": 0.25921144105670874,
+      "grad_norm": 0.9274750807980529,
+      "learning_rate": 8.68572241774678e-06,
+      "loss": 0.0192,
+      "step": 1305
+    },
+    {
+      "epoch": 0.25941007051345716,
+      "grad_norm": 0.6824972255867494,
+      "learning_rate": 8.683547459129211e-06,
+      "loss": 0.0168,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2596086999702056,
+      "grad_norm": 0.6136495451069736,
+      "learning_rate": 8.681370975158328e-06,
+      "loss": 0.0108,
+      "step": 1307
+    },
+    {
+      "epoch": 0.259807329426954,
+      "grad_norm": 0.37108024187396627,
+      "learning_rate": 8.679192966735413e-06,
+      "loss": 0.0109,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2600059588837024,
+      "grad_norm": 0.725413502338458,
+      "learning_rate": 8.677013434762373e-06,
+      "loss": 0.0255,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2602045883404509,
+      "grad_norm": 0.48143907389589863,
+      "learning_rate": 8.674832380141754e-06,
+      "loss": 0.0171,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2604032177971993,
+      "grad_norm": 0.34879379609000266,
+      "learning_rate": 8.672649803776724e-06,
+      "loss": 0.0106,
+      "step": 1311
+    },
+    {
+      "epoch": 0.26060184725394775,
+      "grad_norm": 0.5825209707500266,
+      "learning_rate": 8.670465706571088e-06,
+      "loss": 0.0177,
+      "step": 1312
+    },
+    {
+      "epoch": 0.26080047671069617,
+      "grad_norm": 0.46030714480671636,
+      "learning_rate": 8.66828008942928e-06,
+      "loss": 0.0098,
+      "step": 1313
+    },
+    {
+      "epoch": 0.26099910616744465,
+      "grad_norm": 0.7703888628916056,
+      "learning_rate": 8.66609295325636e-06,
+      "loss": 0.0171,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2611977356241931,
+      "grad_norm": 0.738989269533668,
+      "learning_rate": 8.663904298958018e-06,
+      "loss": 0.0157,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2613963650809415,
+      "grad_norm": 0.6362768126248891,
+      "learning_rate": 8.661714127440578e-06,
+      "loss": 0.0106,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2615949945376899,
+      "grad_norm": 0.5455888714935196,
+      "learning_rate": 8.659522439610983e-06,
+      "loss": 0.0142,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2617936239944384,
+      "grad_norm": 0.23611279379951763,
+      "learning_rate": 8.657329236376811e-06,
+      "loss": 0.0067,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2619922534511868,
+      "grad_norm": 0.31209832142135896,
+      "learning_rate": 8.655134518646264e-06,
+      "loss": 0.0133,
+      "step": 1319
+    },
+    {
+      "epoch": 0.26219088290793524,
+      "grad_norm": 0.2910883987846584,
+      "learning_rate": 8.652938287328174e-06,
+      "loss": 0.0084,
+      "step": 1320
+    },
+    {
+      "epoch": 0.26238951236468366,
+      "grad_norm": 0.46180981742302674,
+      "learning_rate": 8.650740543331997e-06,
+      "loss": 0.0094,
+      "step": 1321
+    },
+    {
+      "epoch": 0.26258814182143214,
+      "grad_norm": 1.0186730371322448,
+      "learning_rate": 8.648541287567817e-06,
+      "loss": 0.0176,
+      "step": 1322
+    },
+    {
+      "epoch": 0.26278677127818056,
+      "grad_norm": 0.7996672389811389,
+      "learning_rate": 8.646340520946343e-06,
+      "loss": 0.0241,
+      "step": 1323
+    },
+    {
+      "epoch": 0.262985400734929,
+      "grad_norm": 0.40791342137010506,
+      "learning_rate": 8.644138244378912e-06,
+      "loss": 0.0109,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2631840301916774,
+      "grad_norm": 0.5936153143977514,
+      "learning_rate": 8.641934458777482e-06,
+      "loss": 0.0118,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2633826596484259,
+      "grad_norm": 0.4056311995150495,
+      "learning_rate": 8.63972916505464e-06,
+      "loss": 0.0081,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2635812891051743,
+      "grad_norm": 0.7376187474937147,
+      "learning_rate": 8.637522364123596e-06,
+      "loss": 0.0114,
+      "step": 1327
+    },
+    {
+      "epoch": 0.26377991856192273,
+      "grad_norm": 0.3672522299105733,
+      "learning_rate": 8.635314056898185e-06,
+      "loss": 0.0089,
+      "step": 1328
+    },
+    {
+      "epoch": 0.26397854801867116,
+      "grad_norm": 0.7185888449952436,
+      "learning_rate": 8.633104244292862e-06,
+      "loss": 0.0204,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2641771774754196,
+      "grad_norm": 0.6555374705786724,
+      "learning_rate": 8.630892927222709e-06,
+      "loss": 0.0153,
+      "step": 1330
+    },
+    {
+      "epoch": 0.26437580693216806,
+      "grad_norm": 0.4167555548956689,
+      "learning_rate": 8.628680106603433e-06,
+      "loss": 0.0149,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2645744363889165,
+      "grad_norm": 0.6242106937112943,
+      "learning_rate": 8.626465783351357e-06,
+      "loss": 0.0184,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2647730658456649,
+      "grad_norm": 0.42946885487707775,
+      "learning_rate": 8.624249958383433e-06,
+      "loss": 0.0126,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2649716953024133,
+      "grad_norm": 0.9955126087048953,
+      "learning_rate": 8.62203263261723e-06,
+      "loss": 0.0123,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2651703247591618,
+      "grad_norm": 0.3025009366483527,
+      "learning_rate": 8.61981380697094e-06,
+      "loss": 0.0106,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2653689542159102,
+      "grad_norm": 0.3621342251640229,
+      "learning_rate": 8.617593482363379e-06,
+      "loss": 0.0089,
+      "step": 1336
+    },
+    {
+      "epoch": 0.26556758367265865,
+      "grad_norm": 0.9828115025424791,
+      "learning_rate": 8.615371659713979e-06,
+      "loss": 0.0202,
+      "step": 1337
+    },
+    {
+      "epoch": 0.26576621312940707,
+      "grad_norm": 1.4452351020519076,
+      "learning_rate": 8.613148339942796e-06,
+      "loss": 0.0197,
+      "step": 1338
+    },
+    {
+      "epoch": 0.26596484258615555,
+      "grad_norm": 0.6375350442126254,
+      "learning_rate": 8.610923523970502e-06,
+      "loss": 0.0165,
+      "step": 1339
+    },
+    {
+      "epoch": 0.26616347204290397,
+      "grad_norm": 0.4494514278658364,
+      "learning_rate": 8.608697212718396e-06,
+      "loss": 0.0095,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2663621014996524,
+      "grad_norm": 0.8983564965166115,
+      "learning_rate": 8.606469407108385e-06,
+      "loss": 0.0165,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2665607309564008,
+      "grad_norm": 0.6516352105484908,
+      "learning_rate": 8.604240108063004e-06,
+      "loss": 0.0148,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2667593604131493,
+      "grad_norm": 0.5267184317342489,
+      "learning_rate": 8.602009316505407e-06,
+      "loss": 0.0124,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2669579898698977,
+      "grad_norm": 0.47381761284604557,
+      "learning_rate": 8.599777033359355e-06,
+      "loss": 0.0101,
+      "step": 1344
+    },
+    {
+      "epoch": 0.26715661932664614,
+      "grad_norm": 1.6669999572861662,
+      "learning_rate": 8.597543259549241e-06,
+      "loss": 0.0245,
+      "step": 1345
+    },
+    {
+      "epoch": 0.26735524878339456,
+      "grad_norm": 0.4832507507850066,
+      "learning_rate": 8.595307996000066e-06,
+      "loss": 0.0134,
+      "step": 1346
+    },
+    {
+      "epoch": 0.26755387824014304,
+      "grad_norm": 0.5156754052203035,
+      "learning_rate": 8.59307124363745e-06,
+      "loss": 0.0258,
+      "step": 1347
+    },
+    {
+      "epoch": 0.26775250769689146,
+      "grad_norm": 0.8536512969498504,
+      "learning_rate": 8.590833003387628e-06,
+      "loss": 0.0258,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2679511371536399,
+      "grad_norm": 0.9428969265848054,
+      "learning_rate": 8.588593276177458e-06,
+      "loss": 0.019,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2681497666103883,
+      "grad_norm": 0.7301869629392804,
+      "learning_rate": 8.586352062934404e-06,
+      "loss": 0.0145,
+      "step": 1350
+    },
+    {
+      "epoch": 0.26834839606713673,
+      "grad_norm": 0.4718543613104909,
+      "learning_rate": 8.584109364586554e-06,
+      "loss": 0.0091,
+      "step": 1351
+    },
+    {
+      "epoch": 0.2685470255238852,
+      "grad_norm": 0.271372733060467,
+      "learning_rate": 8.581865182062606e-06,
+      "loss": 0.0094,
+      "step": 1352
+    },
+    {
+      "epoch": 0.26874565498063363,
+      "grad_norm": 0.42823082027908455,
+      "learning_rate": 8.57961951629187e-06,
+      "loss": 0.0117,
+      "step": 1353
+    },
+    {
+      "epoch": 0.26894428443738205,
+      "grad_norm": 0.572008987695907,
+      "learning_rate": 8.57737236820428e-06,
+      "loss": 0.019,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2691429138941305,
+      "grad_norm": 1.2931428416128332,
+      "learning_rate": 8.575123738730373e-06,
+      "loss": 0.0184,
+      "step": 1355
+    },
+    {
+      "epoch": 0.26934154335087895,
+      "grad_norm": 0.4474034296024569,
+      "learning_rate": 8.572873628801305e-06,
+      "loss": 0.0101,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2695401728076274,
+      "grad_norm": 0.8648867742502259,
+      "learning_rate": 8.570622039348849e-06,
+      "loss": 0.0132,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2697388022643758,
+      "grad_norm": 0.771431838314264,
+      "learning_rate": 8.56836897130538e-06,
+      "loss": 0.0176,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2699374317211242,
+      "grad_norm": 0.44094886621911206,
+      "learning_rate": 8.566114425603892e-06,
+      "loss": 0.0126,
+      "step": 1359
+    },
+    {
+      "epoch": 0.2701360611778727,
+      "grad_norm": 0.6577711334649768,
+      "learning_rate": 8.563858403177994e-06,
+      "loss": 0.0185,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2703346906346211,
+      "grad_norm": 0.39262415058239325,
+      "learning_rate": 8.5616009049619e-06,
+      "loss": 0.0155,
+      "step": 1361
+    },
+    {
+      "epoch": 0.27053332009136954,
+      "grad_norm": 1.140595423731314,
+      "learning_rate": 8.559341931890436e-06,
+      "loss": 0.0251,
+      "step": 1362
+    },
+    {
+      "epoch": 0.27073194954811797,
+      "grad_norm": 0.4989543570417254,
+      "learning_rate": 8.557081484899043e-06,
+      "loss": 0.016,
+      "step": 1363
+    },
+    {
+      "epoch": 0.27093057900486645,
+      "grad_norm": 0.9607995698578123,
+      "learning_rate": 8.55481956492377e-06,
+      "loss": 0.0186,
+      "step": 1364
+    },
+    {
+      "epoch": 0.27112920846161487,
+      "grad_norm": 0.5416529886893404,
+      "learning_rate": 8.552556172901276e-06,
+      "loss": 0.0172,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2713278379183633,
+      "grad_norm": 0.6712403945203463,
+      "learning_rate": 8.550291309768826e-06,
+      "loss": 0.0112,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2715264673751117,
+      "grad_norm": 0.34387714192881513,
+      "learning_rate": 8.548024976464302e-06,
+      "loss": 0.0094,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2717250968318602,
+      "grad_norm": 0.4152557312415485,
+      "learning_rate": 8.545757173926187e-06,
+      "loss": 0.0148,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2719237262886086,
+      "grad_norm": 0.5580718606691937,
+      "learning_rate": 8.543487903093577e-06,
+      "loss": 0.0112,
+      "step": 1369
+    },
+    {
+      "epoch": 0.27212235574535704,
+      "grad_norm": 0.6980069434101902,
+      "learning_rate": 8.541217164906177e-06,
+      "loss": 0.0197,
+      "step": 1370
+    },
+    {
+      "epoch": 0.27232098520210546,
+      "grad_norm": 0.6713064539961933,
+      "learning_rate": 8.538944960304292e-06,
+      "loss": 0.0237,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2725196146588539,
+      "grad_norm": 0.2791585912778628,
+      "learning_rate": 8.536671290228846e-06,
+      "loss": 0.0096,
+      "step": 1372
+    },
+    {
+      "epoch": 0.27271824411560236,
+      "grad_norm": 1.7948572479774836,
+      "learning_rate": 8.534396155621358e-06,
+      "loss": 0.0206,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2729168735723508,
+      "grad_norm": 0.5146160931306909,
+      "learning_rate": 8.532119557423964e-06,
+      "loss": 0.0078,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2731155030290992,
+      "grad_norm": 0.6336391927705046,
+      "learning_rate": 8.529841496579396e-06,
+      "loss": 0.0154,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2733141324858476,
+      "grad_norm": 1.0228381650063942,
+      "learning_rate": 8.527561974031e-06,
+      "loss": 0.0125,
+      "step": 1376
+    },
+    {
+      "epoch": 0.2735127619425961,
+      "grad_norm": 0.5621590918716233,
+      "learning_rate": 8.525280990722723e-06,
+      "loss": 0.0152,
+      "step": 1377
+    },
+    {
+      "epoch": 0.27371139139934453,
+      "grad_norm": 0.5992082774576349,
+      "learning_rate": 8.52299854759912e-06,
+      "loss": 0.015,
+      "step": 1378
+    },
+    {
+      "epoch": 0.27391002085609295,
+      "grad_norm": 0.5427316635640782,
+      "learning_rate": 8.520714645605344e-06,
+      "loss": 0.0217,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2741086503128414,
+      "grad_norm": 0.7839896301325817,
+      "learning_rate": 8.51842928568716e-06,
+      "loss": 0.0117,
+      "step": 1380
+    },
+    {
+      "epoch": 0.27430727976958985,
+      "grad_norm": 0.4131385983667191,
+      "learning_rate": 8.516142468790931e-06,
+      "loss": 0.0159,
+      "step": 1381
+    },
+    {
+      "epoch": 0.2745059092263383,
+      "grad_norm": 0.4081999784122884,
+      "learning_rate": 8.513854195863629e-06,
+      "loss": 0.0099,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2747045386830867,
+      "grad_norm": 0.4673684154571544,
+      "learning_rate": 8.511564467852822e-06,
+      "loss": 0.0077,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2749031681398351,
+      "grad_norm": 0.5713154043757909,
+      "learning_rate": 8.509273285706686e-06,
+      "loss": 0.0195,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2751017975965836,
+      "grad_norm": 0.46380108544256105,
+      "learning_rate": 8.506980650373995e-06,
+      "loss": 0.0142,
+      "step": 1385
+    },
+    {
+      "epoch": 0.275300427053332,
+      "grad_norm": 1.0792577996847905,
+      "learning_rate": 8.50468656280413e-06,
+      "loss": 0.0222,
+      "step": 1386
+    },
+    {
+      "epoch": 0.27549905651008044,
+      "grad_norm": 0.42746699373310537,
+      "learning_rate": 8.50239102394707e-06,
+      "loss": 0.0146,
+      "step": 1387
+    },
+    {
+      "epoch": 0.27569768596682886,
+      "grad_norm": 1.7457011754660738,
+      "learning_rate": 8.500094034753393e-06,
+      "loss": 0.0197,
+      "step": 1388
+    },
+    {
+      "epoch": 0.27589631542357734,
+      "grad_norm": 0.6820989464273655,
+      "learning_rate": 8.49779559617428e-06,
+      "loss": 0.0134,
+      "step": 1389
+    },
+    {
+      "epoch": 0.27609494488032577,
+      "grad_norm": 0.5308996170906365,
+      "learning_rate": 8.495495709161516e-06,
+      "loss": 0.0196,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2762935743370742,
+      "grad_norm": 0.6992766539498697,
+      "learning_rate": 8.49319437466748e-06,
+      "loss": 0.0097,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2764922037938226,
+      "grad_norm": 0.2293267653263177,
+      "learning_rate": 8.49089159364515e-06,
+      "loss": 0.006,
+      "step": 1392
+    },
+    {
+      "epoch": 0.27669083325057103,
+      "grad_norm": 0.32951425806657547,
+      "learning_rate": 8.488587367048105e-06,
+      "loss": 0.0073,
+      "step": 1393
+    },
+    {
+      "epoch": 0.2768894627073195,
+      "grad_norm": 1.6153168594424325,
+      "learning_rate": 8.486281695830527e-06,
+      "loss": 0.0223,
+      "step": 1394
+    },
+    {
+      "epoch": 0.27708809216406793,
+      "grad_norm": 1.864778170921737,
+      "learning_rate": 8.483974580947189e-06,
+      "loss": 0.0329,
+      "step": 1395
+    },
+    {
+      "epoch": 0.27728672162081636,
+      "grad_norm": 1.7739863390597264,
+      "learning_rate": 8.481666023353468e-06,
+      "loss": 0.0263,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2774853510775648,
+      "grad_norm": 0.5128932906013113,
+      "learning_rate": 8.479356024005332e-06,
+      "loss": 0.0113,
+      "step": 1397
+    },
+    {
+      "epoch": 0.27768398053431326,
+      "grad_norm": 0.820357228467481,
+      "learning_rate": 8.47704458385935e-06,
+      "loss": 0.014,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2778826099910617,
+      "grad_norm": 0.6925209483529865,
+      "learning_rate": 8.47473170387269e-06,
+      "loss": 0.0148,
+      "step": 1399
+    },
+    {
+      "epoch": 0.2780812394478101,
+      "grad_norm": 0.5888881159884609,
+      "learning_rate": 8.472417385003109e-06,
+      "loss": 0.019,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2782798689045585,
+      "grad_norm": 0.8199869161478691,
+      "learning_rate": 8.470101628208966e-06,
+      "loss": 0.0135,
+      "step": 1401
+    },
+    {
+      "epoch": 0.278478498361307,
+      "grad_norm": 0.7210167222946081,
+      "learning_rate": 8.467784434449216e-06,
+      "loss": 0.0172,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2786771278180554,
+      "grad_norm": 0.4196353511380168,
+      "learning_rate": 8.465465804683404e-06,
+      "loss": 0.0155,
+      "step": 1403
+    },
+    {
+      "epoch": 0.27887575727480385,
+      "grad_norm": 0.6034473650484623,
+      "learning_rate": 8.463145739871672e-06,
+      "loss": 0.0096,
+      "step": 1404
+    },
+    {
+      "epoch": 0.27907438673155227,
+      "grad_norm": 0.36016231959989076,
+      "learning_rate": 8.460824240974757e-06,
+      "loss": 0.0103,
+      "step": 1405
+    },
+    {
+      "epoch": 0.27927301618830075,
+      "grad_norm": 0.5475792609192718,
+      "learning_rate": 8.458501308953988e-06,
+      "loss": 0.0154,
+      "step": 1406
+    },
+    {
+      "epoch": 0.27947164564504917,
+      "grad_norm": 0.5922387773813316,
+      "learning_rate": 8.456176944771293e-06,
+      "loss": 0.0152,
+      "step": 1407
+    },
+    {
+      "epoch": 0.2796702751017976,
+      "grad_norm": 0.9948119748458698,
+      "learning_rate": 8.453851149389185e-06,
+      "loss": 0.0245,
+      "step": 1408
+    },
+    {
+      "epoch": 0.279868904558546,
+      "grad_norm": 0.8243450247239436,
+      "learning_rate": 8.451523923770776e-06,
+      "loss": 0.013,
+      "step": 1409
+    },
+    {
+      "epoch": 0.2800675340152945,
+      "grad_norm": 0.7318369517598357,
+      "learning_rate": 8.449195268879767e-06,
+      "loss": 0.0133,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2802661634720429,
+      "grad_norm": 0.5419561316890904,
+      "learning_rate": 8.446865185680448e-06,
+      "loss": 0.0093,
+      "step": 1411
+    },
+    {
+      "epoch": 0.28046479292879134,
+      "grad_norm": 0.5265876588995337,
+      "learning_rate": 8.44453367513771e-06,
+      "loss": 0.0171,
+      "step": 1412
+    },
+    {
+      "epoch": 0.28066342238553976,
+      "grad_norm": 0.6859813997920388,
+      "learning_rate": 8.442200738217025e-06,
+      "loss": 0.0095,
+      "step": 1413
+    },
+    {
+      "epoch": 0.2808620518422882,
+      "grad_norm": 0.5649693302253302,
+      "learning_rate": 8.439866375884464e-06,
+      "loss": 0.0212,
+      "step": 1414
+    },
+    {
+      "epoch": 0.28106068129903666,
+      "grad_norm": 0.5862281313195347,
+      "learning_rate": 8.437530589106679e-06,
+      "loss": 0.0221,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2812593107557851,
+      "grad_norm": 0.33080120075065145,
+      "learning_rate": 8.435193378850921e-06,
+      "loss": 0.0117,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2814579402125335,
+      "grad_norm": 0.9234190022508963,
+      "learning_rate": 8.432854746085024e-06,
+      "loss": 0.0173,
+      "step": 1417
+    },
+    {
+      "epoch": 0.28165656966928193,
+      "grad_norm": 0.8430616129766179,
+      "learning_rate": 8.430514691777415e-06,
+      "loss": 0.0145,
+      "step": 1418
+    },
+    {
+      "epoch": 0.2818551991260304,
+      "grad_norm": 0.8475901106471433,
+      "learning_rate": 8.428173216897107e-06,
+      "loss": 0.0159,
+      "step": 1419
+    },
+    {
+      "epoch": 0.28205382858277883,
+      "grad_norm": 0.5735491390921624,
+      "learning_rate": 8.425830322413703e-06,
+      "loss": 0.0097,
+      "step": 1420
+    },
+    {
+      "epoch": 0.28225245803952725,
+      "grad_norm": 0.5266507760548224,
+      "learning_rate": 8.423486009297394e-06,
+      "loss": 0.0109,
+      "step": 1421
+    },
+    {
+      "epoch": 0.2824510874962757,
+      "grad_norm": 0.36869636591820587,
+      "learning_rate": 8.421140278518955e-06,
+      "loss": 0.0085,
+      "step": 1422
+    },
+    {
+      "epoch": 0.28264971695302415,
+      "grad_norm": 1.2202844836952085,
+      "learning_rate": 8.418793131049757e-06,
+      "loss": 0.0266,
+      "step": 1423
+    },
+    {
+      "epoch": 0.2828483464097726,
+      "grad_norm": 0.6504661966699112,
+      "learning_rate": 8.416444567861742e-06,
+      "loss": 0.0106,
+      "step": 1424
+    },
+    {
+      "epoch": 0.283046975866521,
+      "grad_norm": 1.5018651106882912,
+      "learning_rate": 8.414094589927455e-06,
+      "loss": 0.018,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2832456053232694,
+      "grad_norm": 0.48052848756832556,
+      "learning_rate": 8.411743198220016e-06,
+      "loss": 0.0106,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2834442347800179,
+      "grad_norm": 0.7465489964730609,
+      "learning_rate": 8.409390393713139e-06,
+      "loss": 0.0118,
+      "step": 1427
+    },
+    {
+      "epoch": 0.2836428642367663,
+      "grad_norm": 0.49935340474238005,
+      "learning_rate": 8.407036177381111e-06,
+      "loss": 0.0068,
+      "step": 1428
+    },
+    {
+      "epoch": 0.28384149369351475,
+      "grad_norm": 0.6438837817206333,
+      "learning_rate": 8.404680550198814e-06,
+      "loss": 0.0112,
+      "step": 1429
+    },
+    {
+      "epoch": 0.28404012315026317,
+      "grad_norm": 0.46587307611736145,
+      "learning_rate": 8.40232351314171e-06,
+      "loss": 0.0083,
+      "step": 1430
+    },
+    {
+      "epoch": 0.28423875260701165,
+      "grad_norm": 1.2474325824762231,
+      "learning_rate": 8.399965067185849e-06,
+      "loss": 0.0155,
+      "step": 1431
+    },
+    {
+      "epoch": 0.28443738206376007,
+      "grad_norm": 1.0245682468407986,
+      "learning_rate": 8.397605213307858e-06,
+      "loss": 0.0185,
+      "step": 1432
+    },
+    {
+      "epoch": 0.2846360115205085,
+      "grad_norm": 0.7066781397822776,
+      "learning_rate": 8.395243952484949e-06,
+      "loss": 0.0107,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2848346409772569,
+      "grad_norm": 0.7550032030601233,
+      "learning_rate": 8.392881285694918e-06,
+      "loss": 0.015,
+      "step": 1434
+    },
+    {
+      "epoch": 0.28503327043400534,
+      "grad_norm": 0.6874113816132006,
+      "learning_rate": 8.390517213916147e-06,
+      "loss": 0.0142,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2852318998907538,
+      "grad_norm": 1.2976316129854208,
+      "learning_rate": 8.388151738127592e-06,
+      "loss": 0.0181,
+      "step": 1436
+    },
+    {
+      "epoch": 0.28543052934750224,
+      "grad_norm": 1.326879456283606,
+      "learning_rate": 8.385784859308796e-06,
+      "loss": 0.0166,
+      "step": 1437
+    },
+    {
+      "epoch": 0.28562915880425066,
+      "grad_norm": 1.086178549836567,
+      "learning_rate": 8.383416578439881e-06,
+      "loss": 0.0232,
+      "step": 1438
+    },
+    {
+      "epoch": 0.2858277882609991,
+      "grad_norm": 0.6673231654729204,
+      "learning_rate": 8.381046896501547e-06,
+      "loss": 0.0112,
+      "step": 1439
+    },
+    {
+      "epoch": 0.28602641771774756,
+      "grad_norm": 0.530064954612848,
+      "learning_rate": 8.378675814475081e-06,
+      "loss": 0.0062,
+      "step": 1440
+    },
+    {
+      "epoch": 0.286225047174496,
+      "grad_norm": 1.1957721069947402,
+      "learning_rate": 8.376303333342342e-06,
+      "loss": 0.0206,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2864236766312444,
+      "grad_norm": 0.45312372809065454,
+      "learning_rate": 8.373929454085775e-06,
+      "loss": 0.0085,
+      "step": 1442
+    },
+    {
+      "epoch": 0.28662230608799283,
+      "grad_norm": 0.332625950245216,
+      "learning_rate": 8.371554177688399e-06,
+      "loss": 0.0087,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2868209355447413,
+      "grad_norm": 0.6970930696458172,
+      "learning_rate": 8.369177505133814e-06,
+      "loss": 0.0131,
+      "step": 1444
+    },
+    {
+      "epoch": 0.28701956500148973,
+      "grad_norm": 1.1105719027213687,
+      "learning_rate": 8.3667994374062e-06,
+      "loss": 0.0239,
+      "step": 1445
+    },
+    {
+      "epoch": 0.28721819445823815,
+      "grad_norm": 1.0006129734400397,
+      "learning_rate": 8.36441997549031e-06,
+      "loss": 0.0174,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2874168239149866,
+      "grad_norm": 1.4747499745629573,
+      "learning_rate": 8.362039120371475e-06,
+      "loss": 0.0159,
+      "step": 1447
+    },
+    {
+      "epoch": 0.28761545337173505,
+      "grad_norm": 1.0411053353229378,
+      "learning_rate": 8.35965687303561e-06,
+      "loss": 0.0188,
+      "step": 1448
+    },
+    {
+      "epoch": 0.2878140828284835,
+      "grad_norm": 0.9172298750699612,
+      "learning_rate": 8.357273234469196e-06,
+      "loss": 0.0153,
+      "step": 1449
+    },
+    {
+      "epoch": 0.2880127122852319,
+      "grad_norm": 0.3913985500318929,
+      "learning_rate": 8.354888205659299e-06,
+      "loss": 0.0079,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2882113417419803,
+      "grad_norm": 1.374573151737498,
+      "learning_rate": 8.352501787593557e-06,
+      "loss": 0.0281,
+      "step": 1451
+    },
+    {
+      "epoch": 0.2884099711987288,
+      "grad_norm": 0.5566702813433198,
+      "learning_rate": 8.35011398126018e-06,
+      "loss": 0.0173,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2886086006554772,
+      "grad_norm": 1.199039497209081,
+      "learning_rate": 8.347724787647959e-06,
+      "loss": 0.0248,
+      "step": 1453
+    },
+    {
+      "epoch": 0.28880723011222564,
+      "grad_norm": 0.9629869712993756,
+      "learning_rate": 8.345334207746256e-06,
+      "loss": 0.0203,
+      "step": 1454
+    },
+    {
+      "epoch": 0.28900585956897407,
+      "grad_norm": 0.7850010176416995,
+      "learning_rate": 8.342942242545007e-06,
+      "loss": 0.0141,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2892044890257225,
+      "grad_norm": 0.8834250404360167,
+      "learning_rate": 8.340548893034723e-06,
+      "loss": 0.0182,
+      "step": 1456
+    },
+    {
+      "epoch": 0.28940311848247097,
+      "grad_norm": 0.7477391277298652,
+      "learning_rate": 8.338154160206489e-06,
+      "loss": 0.0126,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2896017479392194,
+      "grad_norm": 0.991856579976285,
+      "learning_rate": 8.335758045051959e-06,
+      "loss": 0.0191,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2898003773959678,
+      "grad_norm": 1.5855642011770337,
+      "learning_rate": 8.333360548563363e-06,
+      "loss": 0.0247,
+      "step": 1459
+    },
+    {
+      "epoch": 0.28999900685271623,
+      "grad_norm": 1.3241126782557957,
+      "learning_rate": 8.330961671733503e-06,
+      "loss": 0.0165,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2901976363094647,
+      "grad_norm": 1.4795489122284058,
+      "learning_rate": 8.32856141555575e-06,
+      "loss": 0.024,
+      "step": 1461
+    },
+    {
+      "epoch": 0.29039626576621314,
+      "grad_norm": 0.3260206114711344,
+      "learning_rate": 8.326159781024049e-06,
+      "loss": 0.0108,
+      "step": 1462
+    },
+    {
+      "epoch": 0.29059489522296156,
+      "grad_norm": 0.3190347690618339,
+      "learning_rate": 8.32375676913291e-06,
+      "loss": 0.016,
+      "step": 1463
+    },
+    {
+      "epoch": 0.29079352467971,
+      "grad_norm": 0.5725996431341029,
+      "learning_rate": 8.321352380877426e-06,
+      "loss": 0.017,
+      "step": 1464
+    },
+    {
+      "epoch": 0.29099215413645846,
+      "grad_norm": 0.767026255292494,
+      "learning_rate": 8.318946617253244e-06,
+      "loss": 0.017,
+      "step": 1465
+    },
+    {
+      "epoch": 0.2911907835932069,
+      "grad_norm": 0.5629140122364568,
+      "learning_rate": 8.316539479256594e-06,
+      "loss": 0.015,
+      "step": 1466
+    },
+    {
+      "epoch": 0.2913894130499553,
+      "grad_norm": 0.7358139874860172,
+      "learning_rate": 8.314130967884263e-06,
+      "loss": 0.0129,
+      "step": 1467
+    },
+    {
+      "epoch": 0.2915880425067037,
+      "grad_norm": 0.8209343112526951,
+      "learning_rate": 8.311721084133622e-06,
+      "loss": 0.0246,
+      "step": 1468
+    },
+    {
+      "epoch": 0.2917866719634522,
+      "grad_norm": 0.7719668586694085,
+      "learning_rate": 8.309309829002594e-06,
+      "loss": 0.016,
+      "step": 1469
+    },
+    {
+      "epoch": 0.2919853014202006,
+      "grad_norm": 0.41468903664501805,
+      "learning_rate": 8.30689720348968e-06,
+      "loss": 0.0096,
+      "step": 1470
+    },
+    {
+      "epoch": 0.29218393087694905,
+      "grad_norm": 0.3738228357367781,
+      "learning_rate": 8.304483208593944e-06,
+      "loss": 0.0095,
+      "step": 1471
+    },
+    {
+      "epoch": 0.29238256033369747,
+      "grad_norm": 0.24029576983374756,
+      "learning_rate": 8.302067845315023e-06,
+      "loss": 0.0097,
+      "step": 1472
+    },
+    {
+      "epoch": 0.29258118979044595,
+      "grad_norm": 0.5705732197317682,
+      "learning_rate": 8.299651114653113e-06,
+      "loss": 0.0214,
+      "step": 1473
+    },
+    {
+      "epoch": 0.2927798192471944,
+      "grad_norm": 0.922298441563043,
+      "learning_rate": 8.29723301760898e-06,
+      "loss": 0.0185,
+      "step": 1474
+    },
+    {
+      "epoch": 0.2929784487039428,
+      "grad_norm": 0.8041508059426072,
+      "learning_rate": 8.294813555183959e-06,
+      "loss": 0.0157,
+      "step": 1475
+    },
+    {
+      "epoch": 0.2931770781606912,
+      "grad_norm": 0.8308220892287086,
+      "learning_rate": 8.29239272837994e-06,
+      "loss": 0.0208,
+      "step": 1476
+    },
+    {
+      "epoch": 0.29337570761743964,
+      "grad_norm": 0.41812007821452235,
+      "learning_rate": 8.289970538199391e-06,
+      "loss": 0.016,
+      "step": 1477
+    },
+    {
+      "epoch": 0.2935743370741881,
+      "grad_norm": 0.9237991915282489,
+      "learning_rate": 8.28754698564534e-06,
+      "loss": 0.0138,
+      "step": 1478
+    },
+    {
+      "epoch": 0.29377296653093654,
+      "grad_norm": 0.28751992841811247,
+      "learning_rate": 8.285122071721373e-06,
+      "loss": 0.0107,
+      "step": 1479
+    },
+    {
+      "epoch": 0.29397159598768496,
+      "grad_norm": 0.8535192526289043,
+      "learning_rate": 8.282695797431644e-06,
+      "loss": 0.0254,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2941702254444334,
+      "grad_norm": 0.5712038029199457,
+      "learning_rate": 8.280268163780873e-06,
+      "loss": 0.0075,
+      "step": 1481
+    },
+    {
+      "epoch": 0.29436885490118186,
+      "grad_norm": 0.4432551734371519,
+      "learning_rate": 8.27783917177434e-06,
+      "loss": 0.0095,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2945674843579303,
+      "grad_norm": 0.26503631565417796,
+      "learning_rate": 8.27540882241789e-06,
+      "loss": 0.005,
+      "step": 1483
+    },
+    {
+      "epoch": 0.2947661138146787,
+      "grad_norm": 0.864720411309366,
+      "learning_rate": 8.272977116717925e-06,
+      "loss": 0.0164,
+      "step": 1484
+    },
+    {
+      "epoch": 0.29496474327142713,
+      "grad_norm": 0.6284726332727625,
+      "learning_rate": 8.270544055681415e-06,
+      "loss": 0.0089,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2951633727281756,
+      "grad_norm": 0.508270650687714,
+      "learning_rate": 8.268109640315887e-06,
+      "loss": 0.0172,
+      "step": 1486
+    },
+    {
+      "epoch": 0.29536200218492403,
+      "grad_norm": 0.6269790340357052,
+      "learning_rate": 8.26567387162943e-06,
+      "loss": 0.0179,
+      "step": 1487
+    },
+    {
+      "epoch": 0.29556063164167246,
+      "grad_norm": 0.7297545166439348,
+      "learning_rate": 8.263236750630692e-06,
+      "loss": 0.0219,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2957592610984209,
+      "grad_norm": 0.6898505446448555,
+      "learning_rate": 8.260798278328884e-06,
+      "loss": 0.0108,
+      "step": 1489
+    },
+    {
+      "epoch": 0.29595789055516936,
+      "grad_norm": 0.6708782293191564,
+      "learning_rate": 8.258358455733774e-06,
+      "loss": 0.0084,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2961565200119178,
+      "grad_norm": 0.6386660284055108,
+      "learning_rate": 8.25591728385569e-06,
+      "loss": 0.0113,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2963551494686662,
+      "grad_norm": 1.0471163969424355,
+      "learning_rate": 8.25347476370552e-06,
+      "loss": 0.0187,
+      "step": 1492
+    },
+    {
+      "epoch": 0.2965537789254146,
+      "grad_norm": 0.8613429954652995,
+      "learning_rate": 8.251030896294708e-06,
+      "loss": 0.0146,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2967524083821631,
+      "grad_norm": 0.4725238292843531,
+      "learning_rate": 8.248585682635258e-06,
+      "loss": 0.0162,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2969510378389115,
+      "grad_norm": 0.7080357972606163,
+      "learning_rate": 8.246139123739729e-06,
+      "loss": 0.0277,
+      "step": 1495
+    },
+    {
+      "epoch": 0.29714966729565995,
+      "grad_norm": 0.353145384888274,
+      "learning_rate": 8.243691220621241e-06,
+      "loss": 0.0082,
+      "step": 1496
+    },
+    {
+      "epoch": 0.29734829675240837,
+      "grad_norm": 0.5979605491522597,
+      "learning_rate": 8.241241974293466e-06,
+      "loss": 0.0116,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2975469262091568,
+      "grad_norm": 0.7513238451070279,
+      "learning_rate": 8.238791385770638e-06,
+      "loss": 0.0246,
+      "step": 1498
+    },
+    {
+      "epoch": 0.29774555566590527,
+      "grad_norm": 0.33294751800259487,
+      "learning_rate": 8.236339456067538e-06,
+      "loss": 0.0164,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2979441851226537,
+      "grad_norm": 0.7051678355474877,
+      "learning_rate": 8.233886186199508e-06,
+      "loss": 0.0137,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2981428145794021,
+      "grad_norm": 0.5236577910644429,
+      "learning_rate": 8.231431577182452e-06,
+      "loss": 0.0132,
+      "step": 1501
+    },
+    {
+      "epoch": 0.29834144403615054,
+      "grad_norm": 0.47643786878372224,
+      "learning_rate": 8.228975630032812e-06,
+      "loss": 0.0072,
+      "step": 1502
+    },
+    {
+      "epoch": 0.298540073492899,
+      "grad_norm": 0.5218883992156759,
+      "learning_rate": 8.226518345767598e-06,
+      "loss": 0.0172,
+      "step": 1503
+    },
+    {
+      "epoch": 0.29873870294964744,
+      "grad_norm": 0.4673001168694523,
+      "learning_rate": 8.224059725404369e-06,
+      "loss": 0.0125,
+      "step": 1504
+    },
+    {
+      "epoch": 0.29893733240639586,
+      "grad_norm": 0.34307612038889873,
+      "learning_rate": 8.221599769961235e-06,
+      "loss": 0.0156,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2991359618631443,
+      "grad_norm": 1.3513076887859734,
+      "learning_rate": 8.219138480456864e-06,
+      "loss": 0.0209,
+      "step": 1506
+    },
+    {
+      "epoch": 0.29933459131989276,
+      "grad_norm": 0.532303893440917,
+      "learning_rate": 8.21667585791047e-06,
+      "loss": 0.0181,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2995332207766412,
+      "grad_norm": 0.4536212237431732,
+      "learning_rate": 8.214211903341826e-06,
+      "loss": 0.0139,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2997318502333896,
+      "grad_norm": 0.4663375260088001,
+      "learning_rate": 8.211746617771253e-06,
+      "loss": 0.0106,
+      "step": 1509
+    },
+    {
+      "epoch": 0.29993047969013803,
+      "grad_norm": 0.38114356426993984,
+      "learning_rate": 8.209280002219619e-06,
+      "loss": 0.0157,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3001291091468865,
+      "grad_norm": 0.5218807737983572,
+      "learning_rate": 8.206812057708352e-06,
+      "loss": 0.017,
+      "step": 1511
+    },
+    {
+      "epoch": 0.30032773860363493,
+      "grad_norm": 0.8868790496523004,
+      "learning_rate": 8.204342785259423e-06,
+      "loss": 0.0159,
+      "step": 1512
+    },
+    {
+      "epoch": 0.30052636806038335,
+      "grad_norm": 0.3421070765069182,
+      "learning_rate": 8.201872185895355e-06,
+      "loss": 0.0116,
+      "step": 1513
+    },
+    {
+      "epoch": 0.3007249975171318,
+      "grad_norm": 0.9202280991148633,
+      "learning_rate": 8.19940026063922e-06,
+      "loss": 0.0175,
+      "step": 1514
+    },
+    {
+      "epoch": 0.30092362697388025,
+      "grad_norm": 0.3751885600682715,
+      "learning_rate": 8.196927010514642e-06,
+      "loss": 0.0101,
+      "step": 1515
+    },
+    {
+      "epoch": 0.3011222564306287,
+      "grad_norm": 0.5044522661178389,
+      "learning_rate": 8.194452436545792e-06,
+      "loss": 0.0099,
+      "step": 1516
+    },
+    {
+      "epoch": 0.3013208858873771,
+      "grad_norm": 0.9821726841889677,
+      "learning_rate": 8.191976539757385e-06,
+      "loss": 0.0204,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3015195153441255,
+      "grad_norm": 0.33670074524178356,
+      "learning_rate": 8.18949932117469e-06,
+      "loss": 0.0103,
+      "step": 1518
+    },
+    {
+      "epoch": 0.30171814480087394,
+      "grad_norm": 0.42279805484652505,
+      "learning_rate": 8.18702078182352e-06,
+      "loss": 0.0241,
+      "step": 1519
+    },
+    {
+      "epoch": 0.3019167742576224,
+      "grad_norm": 0.4322955622464677,
+      "learning_rate": 8.184540922730237e-06,
+      "loss": 0.0149,
+      "step": 1520
+    },
+    {
+      "epoch": 0.30211540371437084,
+      "grad_norm": 1.2923583198099244,
+      "learning_rate": 8.182059744921745e-06,
+      "loss": 0.0179,
+      "step": 1521
+    },
+    {
+      "epoch": 0.30231403317111927,
+      "grad_norm": 1.0983654911483387,
+      "learning_rate": 8.179577249425501e-06,
+      "loss": 0.0132,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3025126626278677,
+      "grad_norm": 0.6994601865320401,
+      "learning_rate": 8.177093437269503e-06,
+      "loss": 0.0159,
+      "step": 1523
+    },
+    {
+      "epoch": 0.30271129208461617,
+      "grad_norm": 0.9946803548598048,
+      "learning_rate": 8.174608309482293e-06,
+      "loss": 0.0112,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3029099215413646,
+      "grad_norm": 0.5054848367639058,
+      "learning_rate": 8.17212186709296e-06,
+      "loss": 0.0079,
+      "step": 1525
+    },
+    {
+      "epoch": 0.303108550998113,
+      "grad_norm": 1.1552485716095362,
+      "learning_rate": 8.16963411113114e-06,
+      "loss": 0.0139,
+      "step": 1526
+    },
+    {
+      "epoch": 0.30330718045486144,
+      "grad_norm": 0.6284110896879137,
+      "learning_rate": 8.167145042627007e-06,
+      "loss": 0.011,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3035058099116099,
+      "grad_norm": 0.381511628810998,
+      "learning_rate": 8.164654662611285e-06,
+      "loss": 0.0112,
+      "step": 1528
+    },
+    {
+      "epoch": 0.30370443936835834,
+      "grad_norm": 0.9578113568076267,
+      "learning_rate": 8.162162972115235e-06,
+      "loss": 0.0165,
+      "step": 1529
+    },
+    {
+      "epoch": 0.30390306882510676,
+      "grad_norm": 0.3108981457607181,
+      "learning_rate": 8.159669972170667e-06,
+      "loss": 0.0041,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3041016982818552,
+      "grad_norm": 0.387381343425436,
+      "learning_rate": 8.157175663809926e-06,
+      "loss": 0.0119,
+      "step": 1531
+    },
+    {
+      "epoch": 0.30430032773860366,
+      "grad_norm": 0.7862216295997828,
+      "learning_rate": 8.154680048065905e-06,
+      "loss": 0.0224,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3044989571953521,
+      "grad_norm": 0.9430871632430338,
+      "learning_rate": 8.152183125972036e-06,
+      "loss": 0.0301,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3046975866521005,
+      "grad_norm": 0.532176799345402,
+      "learning_rate": 8.149684898562289e-06,
+      "loss": 0.0102,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3048962161088489,
+      "grad_norm": 1.2585493832335066,
+      "learning_rate": 8.147185366871182e-06,
+      "loss": 0.0168,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3050948455655974,
+      "grad_norm": 0.5332469990576837,
+      "learning_rate": 8.144684531933765e-06,
+      "loss": 0.0187,
+      "step": 1536
+    },
+    {
+      "epoch": 0.30529347502234583,
+      "grad_norm": 0.6578476556556282,
+      "learning_rate": 8.142182394785633e-06,
+      "loss": 0.0098,
+      "step": 1537
+    },
+    {
+      "epoch": 0.30549210447909425,
+      "grad_norm": 0.9710766601800311,
+      "learning_rate": 8.139678956462917e-06,
+      "loss": 0.0147,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3056907339358427,
+      "grad_norm": 0.3280996551077205,
+      "learning_rate": 8.13717421800229e-06,
+      "loss": 0.0114,
+      "step": 1539
+    },
+    {
+      "epoch": 0.3058893633925911,
+      "grad_norm": 0.4189662797300033,
+      "learning_rate": 8.134668180440962e-06,
+      "loss": 0.0121,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3060879928493396,
+      "grad_norm": 0.9921003354040996,
+      "learning_rate": 8.13216084481668e-06,
+      "loss": 0.0136,
+      "step": 1541
+    },
+    {
+      "epoch": 0.306286622306088,
+      "grad_norm": 1.00118711821387,
+      "learning_rate": 8.129652212167725e-06,
+      "loss": 0.0126,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3064852517628364,
+      "grad_norm": 1.1824434262355368,
+      "learning_rate": 8.127142283532926e-06,
+      "loss": 0.0206,
+      "step": 1543
+    },
+    {
+      "epoch": 0.30668388121958484,
+      "grad_norm": 0.3324696515714837,
+      "learning_rate": 8.124631059951638e-06,
+      "loss": 0.0117,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3068825106763333,
+      "grad_norm": 0.2722477636883906,
+      "learning_rate": 8.122118542463758e-06,
+      "loss": 0.0069,
+      "step": 1545
+    },
+    {
+      "epoch": 0.30708114013308174,
+      "grad_norm": 0.38654260321116607,
+      "learning_rate": 8.119604732109716e-06,
+      "loss": 0.0128,
+      "step": 1546
+    },
+    {
+      "epoch": 0.30727976958983017,
+      "grad_norm": 0.4093726871049893,
+      "learning_rate": 8.117089629930479e-06,
+      "loss": 0.0094,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3074783990465786,
+      "grad_norm": 1.471363664471753,
+      "learning_rate": 8.114573236967546e-06,
+      "loss": 0.0154,
+      "step": 1548
+    },
+    {
+      "epoch": 0.30767702850332707,
+      "grad_norm": 0.8945599914230751,
+      "learning_rate": 8.112055554262956e-06,
+      "loss": 0.0225,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3078756579600755,
+      "grad_norm": 0.5070483754063576,
+      "learning_rate": 8.109536582859276e-06,
+      "loss": 0.0139,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3080742874168239,
+      "grad_norm": 0.4124395849167642,
+      "learning_rate": 8.107016323799612e-06,
+      "loss": 0.0157,
+      "step": 1551
+    },
+    {
+      "epoch": 0.30827291687357233,
+      "grad_norm": 0.7163339426933371,
+      "learning_rate": 8.1044947781276e-06,
+      "loss": 0.0188,
+      "step": 1552
+    },
+    {
+      "epoch": 0.3084715463303208,
+      "grad_norm": 0.663951545919708,
+      "learning_rate": 8.10197194688741e-06,
+      "loss": 0.0095,
+      "step": 1553
+    },
+    {
+      "epoch": 0.30867017578706923,
+      "grad_norm": 0.4033057526388188,
+      "learning_rate": 8.099447831123742e-06,
+      "loss": 0.0115,
+      "step": 1554
+    },
+    {
+      "epoch": 0.30886880524381766,
+      "grad_norm": 0.5134664101495625,
+      "learning_rate": 8.09692243188183e-06,
+      "loss": 0.0146,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3090674347005661,
+      "grad_norm": 0.3272561207723999,
+      "learning_rate": 8.094395750207443e-06,
+      "loss": 0.0066,
+      "step": 1556
+    },
+    {
+      "epoch": 0.30926606415731456,
+      "grad_norm": 0.47522492760295715,
+      "learning_rate": 8.091867787146874e-06,
+      "loss": 0.0175,
+      "step": 1557
+    },
+    {
+      "epoch": 0.309464693614063,
+      "grad_norm": 0.28831801397227647,
+      "learning_rate": 8.08933854374695e-06,
+      "loss": 0.0087,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3096633230708114,
+      "grad_norm": 0.728136763673717,
+      "learning_rate": 8.086808021055029e-06,
+      "loss": 0.0146,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3098619525275598,
+      "grad_norm": 0.5645255218117098,
+      "learning_rate": 8.084276220118997e-06,
+      "loss": 0.0131,
+      "step": 1560
+    },
+    {
+      "epoch": 0.31006058198430825,
+      "grad_norm": 0.9114867700044899,
+      "learning_rate": 8.081743141987271e-06,
+      "loss": 0.0172,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3102592114410567,
+      "grad_norm": 0.5128080694665796,
+      "learning_rate": 8.079208787708797e-06,
+      "loss": 0.0181,
+      "step": 1562
+    },
+    {
+      "epoch": 0.31045784089780515,
+      "grad_norm": 0.7452682016764512,
+      "learning_rate": 8.076673158333046e-06,
+      "loss": 0.0131,
+      "step": 1563
+    },
+    {
+      "epoch": 0.31065647035455357,
+      "grad_norm": 0.5224597185446708,
+      "learning_rate": 8.074136254910022e-06,
+      "loss": 0.0172,
+      "step": 1564
+    },
+    {
+      "epoch": 0.310855099811302,
+      "grad_norm": 0.2271594138378722,
+      "learning_rate": 8.071598078490254e-06,
+      "loss": 0.0061,
+      "step": 1565
+    },
+    {
+      "epoch": 0.31105372926805047,
+      "grad_norm": 0.3208897701348875,
+      "learning_rate": 8.069058630124798e-06,
+      "loss": 0.0087,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3112523587247989,
+      "grad_norm": 0.4311676394198718,
+      "learning_rate": 8.066517910865235e-06,
+      "loss": 0.0127,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3114509881815473,
+      "grad_norm": 0.3605691884352593,
+      "learning_rate": 8.063975921763675e-06,
+      "loss": 0.0109,
+      "step": 1568
+    },
+    {
+      "epoch": 0.31164961763829574,
+      "grad_norm": 0.5511650816451228,
+      "learning_rate": 8.061432663872757e-06,
+      "loss": 0.014,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3118482470950442,
+      "grad_norm": 0.4705883402855219,
+      "learning_rate": 8.058888138245639e-06,
+      "loss": 0.018,
+      "step": 1570
+    },
+    {
+      "epoch": 0.31204687655179264,
+      "grad_norm": 0.427333107085936,
+      "learning_rate": 8.056342345936005e-06,
+      "loss": 0.013,
+      "step": 1571
+    },
+    {
+      "epoch": 0.31224550600854106,
+      "grad_norm": 0.9899499461011341,
+      "learning_rate": 8.053795287998065e-06,
+      "loss": 0.0104,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3124441354652895,
+      "grad_norm": 0.4747912372785505,
+      "learning_rate": 8.051246965486557e-06,
+      "loss": 0.0073,
+      "step": 1573
+    },
+    {
+      "epoch": 0.31264276492203796,
+      "grad_norm": 0.441160461346295,
+      "learning_rate": 8.048697379456733e-06,
+      "loss": 0.0159,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3128413943787864,
+      "grad_norm": 0.34014297933534443,
+      "learning_rate": 8.04614653096438e-06,
+      "loss": 0.0097,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3130400238355348,
+      "grad_norm": 1.2792886913012314,
+      "learning_rate": 8.043594421065796e-06,
+      "loss": 0.017,
+      "step": 1576
+    },
+    {
+      "epoch": 0.31323865329228323,
+      "grad_norm": 0.6529038511786388,
+      "learning_rate": 8.041041050817813e-06,
+      "loss": 0.0155,
+      "step": 1577
+    },
+    {
+      "epoch": 0.31343728274903165,
+      "grad_norm": 0.4600809221001575,
+      "learning_rate": 8.038486421277775e-06,
+      "loss": 0.0093,
+      "step": 1578
+    },
+    {
+      "epoch": 0.31363591220578013,
+      "grad_norm": 0.5257994692013639,
+      "learning_rate": 8.035930533503554e-06,
+      "loss": 0.012,
+      "step": 1579
+    },
+    {
+      "epoch": 0.31383454166252855,
+      "grad_norm": 1.482595268996986,
+      "learning_rate": 8.033373388553538e-06,
+      "loss": 0.0126,
+      "step": 1580
+    },
+    {
+      "epoch": 0.314033171119277,
+      "grad_norm": 0.697904773183433,
+      "learning_rate": 8.030814987486639e-06,
+      "loss": 0.0129,
+      "step": 1581
+    },
+    {
+      "epoch": 0.3142318005760254,
+      "grad_norm": 1.1437565550456443,
+      "learning_rate": 8.028255331362292e-06,
+      "loss": 0.0153,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3144304300327739,
+      "grad_norm": 0.7568186293758398,
+      "learning_rate": 8.025694421240442e-06,
+      "loss": 0.0098,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3146290594895223,
+      "grad_norm": 1.2713687932469087,
+      "learning_rate": 8.023132258181563e-06,
+      "loss": 0.0233,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3148276889462707,
+      "grad_norm": 0.6183628181054079,
+      "learning_rate": 8.020568843246642e-06,
+      "loss": 0.0174,
+      "step": 1585
+    },
+    {
+      "epoch": 0.31502631840301915,
+      "grad_norm": 0.38716038710078116,
+      "learning_rate": 8.01800417749719e-06,
+      "loss": 0.0072,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3152249478597676,
+      "grad_norm": 0.43109273141996973,
+      "learning_rate": 8.015438261995229e-06,
+      "loss": 0.0085,
+      "step": 1587
+    },
+    {
+      "epoch": 0.31542357731651605,
+      "grad_norm": 0.72428449579244,
+      "learning_rate": 8.012871097803303e-06,
+      "loss": 0.0134,
+      "step": 1588
+    },
+    {
+      "epoch": 0.31562220677326447,
+      "grad_norm": 0.4444888127096748,
+      "learning_rate": 8.010302685984473e-06,
+      "loss": 0.0118,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3158208362300129,
+      "grad_norm": 0.5112452513036903,
+      "learning_rate": 8.007733027602315e-06,
+      "loss": 0.0085,
+      "step": 1590
+    },
+    {
+      "epoch": 0.31601946568676137,
+      "grad_norm": 0.5630116305420747,
+      "learning_rate": 8.005162123720924e-06,
+      "loss": 0.0096,
+      "step": 1591
+    },
+    {
+      "epoch": 0.3162180951435098,
+      "grad_norm": 0.7437776670920336,
+      "learning_rate": 8.002589975404907e-06,
+      "loss": 0.0144,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3164167246002582,
+      "grad_norm": 0.970854724356747,
+      "learning_rate": 8.000016583719386e-06,
+      "loss": 0.0215,
+      "step": 1593
+    },
+    {
+      "epoch": 0.31661535405700664,
+      "grad_norm": 0.8828896563831873,
+      "learning_rate": 7.997441949730003e-06,
+      "loss": 0.0133,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3168139835137551,
+      "grad_norm": 0.6160271865646878,
+      "learning_rate": 7.994866074502911e-06,
+      "loss": 0.0135,
+      "step": 1595
+    },
+    {
+      "epoch": 0.31701261297050354,
+      "grad_norm": 0.5364322361987068,
+      "learning_rate": 7.992288959104776e-06,
+      "loss": 0.0169,
+      "step": 1596
+    },
+    {
+      "epoch": 0.31721124242725196,
+      "grad_norm": 0.46568042207568855,
+      "learning_rate": 7.98971060460278e-06,
+      "loss": 0.0122,
+      "step": 1597
+    },
+    {
+      "epoch": 0.3174098718840004,
+      "grad_norm": 0.5264630544016791,
+      "learning_rate": 7.987131012064615e-06,
+      "loss": 0.0129,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3176085013407488,
+      "grad_norm": 0.6009801290738634,
+      "learning_rate": 7.984550182558492e-06,
+      "loss": 0.0112,
+      "step": 1599
+    },
+    {
+      "epoch": 0.3178071307974973,
+      "grad_norm": 0.6884006171260445,
+      "learning_rate": 7.981968117153125e-06,
+      "loss": 0.0191,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3180057602542457,
+      "grad_norm": 0.3982485689190538,
+      "learning_rate": 7.979384816917748e-06,
+      "loss": 0.0111,
+      "step": 1601
+    },
+    {
+      "epoch": 0.31820438971099413,
+      "grad_norm": 0.30136377831303135,
+      "learning_rate": 7.9768002829221e-06,
+      "loss": 0.0113,
+      "step": 1602
+    },
+    {
+      "epoch": 0.31840301916774255,
+      "grad_norm": 0.8655679733419154,
+      "learning_rate": 7.974214516236438e-06,
+      "loss": 0.0196,
+      "step": 1603
+    },
+    {
+      "epoch": 0.31860164862449103,
+      "grad_norm": 0.8439790204628401,
+      "learning_rate": 7.971627517931523e-06,
+      "loss": 0.0159,
+      "step": 1604
+    },
+    {
+      "epoch": 0.31880027808123945,
+      "grad_norm": 0.6162008342884321,
+      "learning_rate": 7.96903928907863e-06,
+      "loss": 0.0172,
+      "step": 1605
+    },
+    {
+      "epoch": 0.3189989075379879,
+      "grad_norm": 0.2737773167439073,
+      "learning_rate": 7.966449830749538e-06,
+      "loss": 0.0172,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3191975369947363,
+      "grad_norm": 0.505600845809124,
+      "learning_rate": 7.963859144016544e-06,
+      "loss": 0.0164,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3193961664514848,
+      "grad_norm": 0.796834862743995,
+      "learning_rate": 7.961267229952444e-06,
+      "loss": 0.011,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3195947959082332,
+      "grad_norm": 1.2150625750472108,
+      "learning_rate": 7.958674089630551e-06,
+      "loss": 0.0147,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3197934253649816,
+      "grad_norm": 0.3751880775430107,
+      "learning_rate": 7.956079724124681e-06,
+      "loss": 0.0131,
+      "step": 1610
+    },
+    {
+      "epoch": 0.31999205482173004,
+      "grad_norm": 0.2392680116727292,
+      "learning_rate": 7.953484134509158e-06,
+      "loss": 0.0075,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3201906842784785,
+      "grad_norm": 0.3297290154412873,
+      "learning_rate": 7.950887321858811e-06,
+      "loss": 0.0084,
+      "step": 1612
+    },
+    {
+      "epoch": 0.32038931373522694,
+      "grad_norm": 0.6362428644715254,
+      "learning_rate": 7.948289287248979e-06,
+      "loss": 0.0113,
+      "step": 1613
+    },
+    {
+      "epoch": 0.32058794319197537,
+      "grad_norm": 0.703329270973474,
+      "learning_rate": 7.945690031755506e-06,
+      "loss": 0.0121,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3207865726487238,
+      "grad_norm": 1.1210492893128057,
+      "learning_rate": 7.943089556454742e-06,
+      "loss": 0.0161,
+      "step": 1615
+    },
+    {
+      "epoch": 0.32098520210547227,
+      "grad_norm": 0.7398560335941972,
+      "learning_rate": 7.940487862423538e-06,
+      "loss": 0.0153,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3211838315622207,
+      "grad_norm": 0.592690898155172,
+      "learning_rate": 7.937884950739255e-06,
+      "loss": 0.0126,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3213824610189691,
+      "grad_norm": 0.3885094064525805,
+      "learning_rate": 7.935280822479758e-06,
+      "loss": 0.0126,
+      "step": 1618
+    },
+    {
+      "epoch": 0.32158109047571753,
+      "grad_norm": 0.37641651576137636,
+      "learning_rate": 7.93267547872341e-06,
+      "loss": 0.0158,
+      "step": 1619
+    },
+    {
+      "epoch": 0.32177971993246596,
+      "grad_norm": 0.5131102552628134,
+      "learning_rate": 7.930068920549084e-06,
+      "loss": 0.0077,
+      "step": 1620
+    },
+    {
+      "epoch": 0.32197834938921444,
+      "grad_norm": 0.6672345424407569,
+      "learning_rate": 7.927461149036153e-06,
+      "loss": 0.0154,
+      "step": 1621
+    },
+    {
+      "epoch": 0.32217697884596286,
+      "grad_norm": 0.7516794337115568,
+      "learning_rate": 7.924852165264491e-06,
+      "loss": 0.0119,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3223756083027113,
+      "grad_norm": 0.3990098639952792,
+      "learning_rate": 7.922241970314475e-06,
+      "loss": 0.0089,
+      "step": 1623
+    },
+    {
+      "epoch": 0.3225742377594597,
+      "grad_norm": 0.48114802980288773,
+      "learning_rate": 7.919630565266987e-06,
+      "loss": 0.0123,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3227728672162082,
+      "grad_norm": 0.5316138344004034,
+      "learning_rate": 7.917017951203406e-06,
+      "loss": 0.0064,
+      "step": 1625
+    },
+    {
+      "epoch": 0.3229714966729566,
+      "grad_norm": 0.30950798522402956,
+      "learning_rate": 7.91440412920561e-06,
+      "loss": 0.004,
+      "step": 1626
+    },
+    {
+      "epoch": 0.323170126129705,
+      "grad_norm": 0.8119698454248738,
+      "learning_rate": 7.911789100355985e-06,
+      "loss": 0.017,
+      "step": 1627
+    },
+    {
+      "epoch": 0.32336875558645345,
+      "grad_norm": 0.3266822113853905,
+      "learning_rate": 7.90917286573741e-06,
+      "loss": 0.0112,
+      "step": 1628
+    },
+    {
+      "epoch": 0.3235673850432019,
+      "grad_norm": 0.5213109216327177,
+      "learning_rate": 7.906555426433264e-06,
+      "loss": 0.0084,
+      "step": 1629
+    },
+    {
+      "epoch": 0.32376601449995035,
+      "grad_norm": 1.0671826630070904,
+      "learning_rate": 7.903936783527425e-06,
+      "loss": 0.0198,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3239646439566988,
+      "grad_norm": 1.0574028610546085,
+      "learning_rate": 7.901316938104275e-06,
+      "loss": 0.0148,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3241632734134472,
+      "grad_norm": 0.46764434451459613,
+      "learning_rate": 7.898695891248685e-06,
+      "loss": 0.0064,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3243619028701957,
+      "grad_norm": 0.8539693335630968,
+      "learning_rate": 7.896073644046028e-06,
+      "loss": 0.0135,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3245605323269441,
+      "grad_norm": 0.8064389257572152,
+      "learning_rate": 7.893450197582178e-06,
+      "loss": 0.0117,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3247591617836925,
+      "grad_norm": 0.34763349103097907,
+      "learning_rate": 7.890825552943495e-06,
+      "loss": 0.0075,
+      "step": 1635
+    },
+    {
+      "epoch": 0.32495779124044094,
+      "grad_norm": 0.48244504223164625,
+      "learning_rate": 7.888199711216848e-06,
+      "loss": 0.0138,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3251564206971894,
+      "grad_norm": 0.5896875038752724,
+      "learning_rate": 7.885572673489592e-06,
+      "loss": 0.0126,
+      "step": 1637
+    },
+    {
+      "epoch": 0.32535505015393784,
+      "grad_norm": 0.6571748595454835,
+      "learning_rate": 7.882944440849582e-06,
+      "loss": 0.0117,
+      "step": 1638
+    },
+    {
+      "epoch": 0.32555367961068626,
+      "grad_norm": 0.31453304815725347,
+      "learning_rate": 7.880315014385166e-06,
+      "loss": 0.0048,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3257523090674347,
+      "grad_norm": 1.1519220230010478,
+      "learning_rate": 7.877684395185187e-06,
+      "loss": 0.0127,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3259509385241831,
+      "grad_norm": 0.770459673164255,
+      "learning_rate": 7.875052584338983e-06,
+      "loss": 0.0206,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3261495679809316,
+      "grad_norm": 0.7015223884365756,
+      "learning_rate": 7.872419582936382e-06,
+      "loss": 0.0175,
+      "step": 1642
+    },
+    {
+      "epoch": 0.32634819743768,
+      "grad_norm": 0.6126064018057952,
+      "learning_rate": 7.869785392067707e-06,
+      "loss": 0.0171,
+      "step": 1643
+    },
+    {
+      "epoch": 0.32654682689442843,
+      "grad_norm": 0.43187448867276834,
+      "learning_rate": 7.867150012823777e-06,
+      "loss": 0.0134,
+      "step": 1644
+    },
+    {
+      "epoch": 0.32674545635117686,
+      "grad_norm": 0.9447876200418266,
+      "learning_rate": 7.864513446295896e-06,
+      "loss": 0.0143,
+      "step": 1645
+    },
+    {
+      "epoch": 0.32694408580792533,
+      "grad_norm": 0.5123267111128992,
+      "learning_rate": 7.861875693575866e-06,
+      "loss": 0.0092,
+      "step": 1646
+    },
+    {
+      "epoch": 0.32714271526467376,
+      "grad_norm": 0.8471560674383536,
+      "learning_rate": 7.859236755755978e-06,
+      "loss": 0.0166,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3273413447214222,
+      "grad_norm": 1.1733673039140131,
+      "learning_rate": 7.856596633929012e-06,
+      "loss": 0.0225,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3275399741781706,
+      "grad_norm": 0.5145085603141826,
+      "learning_rate": 7.85395532918824e-06,
+      "loss": 0.0169,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3277386036349191,
+      "grad_norm": 0.4349990032596344,
+      "learning_rate": 7.851312842627426e-06,
+      "loss": 0.0179,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3279372330916675,
+      "grad_norm": 0.6127954085037567,
+      "learning_rate": 7.848669175340818e-06,
+      "loss": 0.0163,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3281358625484159,
+      "grad_norm": 0.6906096363587637,
+      "learning_rate": 7.846024328423157e-06,
+      "loss": 0.0136,
+      "step": 1652
+    },
+    {
+      "epoch": 0.32833449200516435,
+      "grad_norm": 0.3831242188980579,
+      "learning_rate": 7.843378302969674e-06,
+      "loss": 0.0115,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3285331214619128,
+      "grad_norm": 0.31854883023090935,
+      "learning_rate": 7.840731100076081e-06,
+      "loss": 0.0119,
+      "step": 1654
+    },
+    {
+      "epoch": 0.32873175091866125,
+      "grad_norm": 0.6135826551661053,
+      "learning_rate": 7.838082720838585e-06,
+      "loss": 0.0116,
+      "step": 1655
+    },
+    {
+      "epoch": 0.32893038037540967,
+      "grad_norm": 0.6788478557117993,
+      "learning_rate": 7.835433166353876e-06,
+      "loss": 0.0196,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3291290098321581,
+      "grad_norm": 0.5204341679322476,
+      "learning_rate": 7.832782437719132e-06,
+      "loss": 0.0176,
+      "step": 1657
+    },
+    {
+      "epoch": 0.32932763928890657,
+      "grad_norm": 0.7020429742008044,
+      "learning_rate": 7.830130536032017e-06,
+      "loss": 0.0101,
+      "step": 1658
+    },
+    {
+      "epoch": 0.329526268745655,
+      "grad_norm": 0.782017865564206,
+      "learning_rate": 7.827477462390683e-06,
+      "loss": 0.0156,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3297248982024034,
+      "grad_norm": 0.7872235943247256,
+      "learning_rate": 7.824823217893762e-06,
+      "loss": 0.0183,
+      "step": 1660
+    },
+    {
+      "epoch": 0.32992352765915184,
+      "grad_norm": 0.9577077977386234,
+      "learning_rate": 7.822167803640375e-06,
+      "loss": 0.017,
+      "step": 1661
+    },
+    {
+      "epoch": 0.33012215711590026,
+      "grad_norm": 0.3392933900875892,
+      "learning_rate": 7.819511220730127e-06,
+      "loss": 0.0069,
+      "step": 1662
+    },
+    {
+      "epoch": 0.33032078657264874,
+      "grad_norm": 0.769515283319616,
+      "learning_rate": 7.816853470263107e-06,
+      "loss": 0.0132,
+      "step": 1663
+    },
+    {
+      "epoch": 0.33051941602939716,
+      "grad_norm": 0.41436036382418884,
+      "learning_rate": 7.814194553339884e-06,
+      "loss": 0.0093,
+      "step": 1664
+    },
+    {
+      "epoch": 0.3307180454861456,
+      "grad_norm": 0.7489421767929256,
+      "learning_rate": 7.811534471061516e-06,
+      "loss": 0.015,
+      "step": 1665
+    },
+    {
+      "epoch": 0.330916674942894,
+      "grad_norm": 0.5296856540549193,
+      "learning_rate": 7.80887322452954e-06,
+      "loss": 0.0173,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3311153043996425,
+      "grad_norm": 0.5649510594604672,
+      "learning_rate": 7.806210814845974e-06,
+      "loss": 0.0102,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3313139338563909,
+      "grad_norm": 0.4243877853275429,
+      "learning_rate": 7.803547243113319e-06,
+      "loss": 0.0108,
+      "step": 1668
+    },
+    {
+      "epoch": 0.33151256331313933,
+      "grad_norm": 0.6981372474644114,
+      "learning_rate": 7.800882510434559e-06,
+      "loss": 0.013,
+      "step": 1669
+    },
+    {
+      "epoch": 0.33171119276988775,
+      "grad_norm": 0.5550131373261892,
+      "learning_rate": 7.798216617913155e-06,
+      "loss": 0.0213,
+      "step": 1670
+    },
+    {
+      "epoch": 0.33190982222663623,
+      "grad_norm": 0.867353442467047,
+      "learning_rate": 7.795549566653054e-06,
+      "loss": 0.0116,
+      "step": 1671
+    },
+    {
+      "epoch": 0.33210845168338465,
+      "grad_norm": 0.587120077251956,
+      "learning_rate": 7.792881357758674e-06,
+      "loss": 0.0177,
+      "step": 1672
+    },
+    {
+      "epoch": 0.3323070811401331,
+      "grad_norm": 0.5107619089303768,
+      "learning_rate": 7.790211992334923e-06,
+      "loss": 0.0184,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3325057105968815,
+      "grad_norm": 0.9529228340714601,
+      "learning_rate": 7.787541471487178e-06,
+      "loss": 0.0116,
+      "step": 1674
+    },
+    {
+      "epoch": 0.33270434005363,
+      "grad_norm": 0.32762475045703154,
+      "learning_rate": 7.784869796321302e-06,
+      "loss": 0.0111,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3329029695103784,
+      "grad_norm": 1.0160627610759643,
+      "learning_rate": 7.782196967943633e-06,
+      "loss": 0.0174,
+      "step": 1676
+    },
+    {
+      "epoch": 0.3331015989671268,
+      "grad_norm": 0.8755877388840554,
+      "learning_rate": 7.779522987460985e-06,
+      "loss": 0.0241,
+      "step": 1677
+    },
+    {
+      "epoch": 0.33330022842387524,
+      "grad_norm": 0.4517222646647724,
+      "learning_rate": 7.776847855980653e-06,
+      "loss": 0.0105,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3334988578806237,
+      "grad_norm": 0.4723207594634246,
+      "learning_rate": 7.774171574610404e-06,
+      "loss": 0.0156,
+      "step": 1679
+    },
+    {
+      "epoch": 0.33369748733737215,
+      "grad_norm": 0.5220346418239198,
+      "learning_rate": 7.771494144458483e-06,
+      "loss": 0.0116,
+      "step": 1680
+    },
+    {
+      "epoch": 0.33389611679412057,
+      "grad_norm": 0.2821678482563408,
+      "learning_rate": 7.768815566633612e-06,
+      "loss": 0.0082,
+      "step": 1681
+    },
+    {
+      "epoch": 0.334094746250869,
+      "grad_norm": 1.42187836812918,
+      "learning_rate": 7.766135842244988e-06,
+      "loss": 0.0241,
+      "step": 1682
+    },
+    {
+      "epoch": 0.3342933757076174,
+      "grad_norm": 0.6174532778541035,
+      "learning_rate": 7.763454972402282e-06,
+      "loss": 0.0128,
+      "step": 1683
+    },
+    {
+      "epoch": 0.3344920051643659,
+      "grad_norm": 0.5883317240537213,
+      "learning_rate": 7.76077295821564e-06,
+      "loss": 0.0144,
+      "step": 1684
+    },
+    {
+      "epoch": 0.3346906346211143,
+      "grad_norm": 0.3588975768808438,
+      "learning_rate": 7.75808980079568e-06,
+      "loss": 0.0097,
+      "step": 1685
+    },
+    {
+      "epoch": 0.33488926407786274,
+      "grad_norm": 0.5875630793918583,
+      "learning_rate": 7.755405501253496e-06,
+      "loss": 0.0157,
+      "step": 1686
+    },
+    {
+      "epoch": 0.33508789353461116,
+      "grad_norm": 0.7281884043493897,
+      "learning_rate": 7.752720060700652e-06,
+      "loss": 0.02,
+      "step": 1687
+    },
+    {
+      "epoch": 0.33528652299135964,
+      "grad_norm": 0.6001706983891172,
+      "learning_rate": 7.750033480249188e-06,
+      "loss": 0.0155,
+      "step": 1688
+    },
+    {
+      "epoch": 0.33548515244810806,
+      "grad_norm": 0.2997842597354729,
+      "learning_rate": 7.747345761011616e-06,
+      "loss": 0.0135,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3356837819048565,
+      "grad_norm": 0.5174410489177036,
+      "learning_rate": 7.744656904100913e-06,
+      "loss": 0.0093,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3358824113616049,
+      "grad_norm": 0.5480320708273444,
+      "learning_rate": 7.741966910630536e-06,
+      "loss": 0.0172,
+      "step": 1691
+    },
+    {
+      "epoch": 0.3360810408183534,
+      "grad_norm": 0.4344915109534724,
+      "learning_rate": 7.739275781714405e-06,
+      "loss": 0.0129,
+      "step": 1692
+    },
+    {
+      "epoch": 0.3362796702751018,
+      "grad_norm": 0.8908712492790886,
+      "learning_rate": 7.736583518466919e-06,
+      "loss": 0.0127,
+      "step": 1693
+    },
+    {
+      "epoch": 0.33647829973185023,
+      "grad_norm": 0.849787263342614,
+      "learning_rate": 7.733890122002936e-06,
+      "loss": 0.0119,
+      "step": 1694
+    },
+    {
+      "epoch": 0.33667692918859865,
+      "grad_norm": 0.5855213279534514,
+      "learning_rate": 7.731195593437793e-06,
+      "loss": 0.012,
+      "step": 1695
+    },
+    {
+      "epoch": 0.33687555864534713,
+      "grad_norm": 0.6288057804890751,
+      "learning_rate": 7.728499933887288e-06,
+      "loss": 0.0142,
+      "step": 1696
+    },
+    {
+      "epoch": 0.33707418810209555,
+      "grad_norm": 0.43026765663973393,
+      "learning_rate": 7.725803144467695e-06,
+      "loss": 0.0116,
+      "step": 1697
+    },
+    {
+      "epoch": 0.337272817558844,
+      "grad_norm": 0.36782345928729954,
+      "learning_rate": 7.723105226295749e-06,
+      "loss": 0.0086,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3374714470155924,
+      "grad_norm": 0.9472170162174446,
+      "learning_rate": 7.720406180488655e-06,
+      "loss": 0.0262,
+      "step": 1699
+    },
+    {
+      "epoch": 0.3376700764723409,
+      "grad_norm": 0.3625055211307673,
+      "learning_rate": 7.717706008164085e-06,
+      "loss": 0.0079,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3378687059290893,
+      "grad_norm": 0.2153089494894364,
+      "learning_rate": 7.715004710440181e-06,
+      "loss": 0.0045,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3380673353858377,
+      "grad_norm": 1.055365098303852,
+      "learning_rate": 7.712302288435545e-06,
+      "loss": 0.0146,
+      "step": 1702
+    },
+    {
+      "epoch": 0.33826596484258614,
+      "grad_norm": 1.593570003099904,
+      "learning_rate": 7.709598743269246e-06,
+      "loss": 0.0264,
+      "step": 1703
+    },
+    {
+      "epoch": 0.33846459429933456,
+      "grad_norm": 0.8049164644185485,
+      "learning_rate": 7.70689407606082e-06,
+      "loss": 0.0127,
+      "step": 1704
+    },
+    {
+      "epoch": 0.33866322375608304,
+      "grad_norm": 1.02619924035175,
+      "learning_rate": 7.70418828793027e-06,
+      "loss": 0.0222,
+      "step": 1705
+    },
+    {
+      "epoch": 0.33886185321283147,
+      "grad_norm": 0.3728588121113957,
+      "learning_rate": 7.701481379998057e-06,
+      "loss": 0.0108,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3390604826695799,
+      "grad_norm": 0.6574219495167848,
+      "learning_rate": 7.698773353385111e-06,
+      "loss": 0.0121,
+      "step": 1707
+    },
+    {
+      "epoch": 0.3392591121263283,
+      "grad_norm": 0.38503544600597717,
+      "learning_rate": 7.696064209212822e-06,
+      "loss": 0.0112,
+      "step": 1708
+    },
+    {
+      "epoch": 0.3394577415830768,
+      "grad_norm": 0.558349825474669,
+      "learning_rate": 7.693353948603041e-06,
+      "loss": 0.008,
+      "step": 1709
+    },
+    {
+      "epoch": 0.3396563710398252,
+      "grad_norm": 2.1539383186369774,
+      "learning_rate": 7.69064257267809e-06,
+      "loss": 0.0121,
+      "step": 1710
+    },
+    {
+      "epoch": 0.33985500049657363,
+      "grad_norm": 0.41496259994783097,
+      "learning_rate": 7.687930082560744e-06,
+      "loss": 0.0127,
+      "step": 1711
+    },
+    {
+      "epoch": 0.34005362995332206,
+      "grad_norm": 0.48924590860014766,
+      "learning_rate": 7.685216479374242e-06,
+      "loss": 0.0137,
+      "step": 1712
+    },
+    {
+      "epoch": 0.34025225941007053,
+      "grad_norm": 0.45250983111826154,
+      "learning_rate": 7.682501764242284e-06,
+      "loss": 0.0089,
+      "step": 1713
+    },
+    {
+      "epoch": 0.34045088886681896,
+      "grad_norm": 0.45711497918986704,
+      "learning_rate": 7.679785938289032e-06,
+      "loss": 0.0132,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3406495183235674,
+      "grad_norm": 0.44128917687358954,
+      "learning_rate": 7.677069002639109e-06,
+      "loss": 0.014,
+      "step": 1715
+    },
+    {
+      "epoch": 0.3408481477803158,
+      "grad_norm": 0.7869148221018819,
+      "learning_rate": 7.674350958417589e-06,
+      "loss": 0.018,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3410467772370643,
+      "grad_norm": 0.4703664064866459,
+      "learning_rate": 7.671631806750018e-06,
+      "loss": 0.0144,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3412454066938127,
+      "grad_norm": 0.42705552055783985,
+      "learning_rate": 7.66891154876239e-06,
+      "loss": 0.0067,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3414440361505611,
+      "grad_norm": 1.4239786275656603,
+      "learning_rate": 7.666190185581164e-06,
+      "loss": 0.0215,
+      "step": 1719
+    },
+    {
+      "epoch": 0.34164266560730955,
+      "grad_norm": 0.8964200804878224,
+      "learning_rate": 7.66346771833325e-06,
+      "loss": 0.0105,
+      "step": 1720
+    },
+    {
+      "epoch": 0.341841295064058,
+      "grad_norm": 0.3234278375891176,
+      "learning_rate": 7.660744148146022e-06,
+      "loss": 0.0204,
+      "step": 1721
+    },
+    {
+      "epoch": 0.34203992452080645,
+      "grad_norm": 0.6931183761335391,
+      "learning_rate": 7.658019476147307e-06,
+      "loss": 0.0201,
+      "step": 1722
+    },
+    {
+      "epoch": 0.34223855397755487,
+      "grad_norm": 0.7400781197903014,
+      "learning_rate": 7.65529370346539e-06,
+      "loss": 0.0164,
+      "step": 1723
+    },
+    {
+      "epoch": 0.3424371834343033,
+      "grad_norm": 0.4335482304440916,
+      "learning_rate": 7.652566831229007e-06,
+      "loss": 0.0091,
+      "step": 1724
+    },
+    {
+      "epoch": 0.3426358128910517,
+      "grad_norm": 0.1969898697189572,
+      "learning_rate": 7.649838860567356e-06,
+      "loss": 0.0038,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3428344423478002,
+      "grad_norm": 0.6032444248728773,
+      "learning_rate": 7.647109792610087e-06,
+      "loss": 0.0094,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3430330718045486,
+      "grad_norm": 0.4008378028364625,
+      "learning_rate": 7.644379628487305e-06,
+      "loss": 0.0116,
+      "step": 1727
+    },
+    {
+      "epoch": 0.34323170126129704,
+      "grad_norm": 0.45794703272131426,
+      "learning_rate": 7.641648369329566e-06,
+      "loss": 0.0081,
+      "step": 1728
+    },
+    {
+      "epoch": 0.34343033071804546,
+      "grad_norm": 0.8915628644941473,
+      "learning_rate": 7.638916016267884e-06,
+      "loss": 0.0136,
+      "step": 1729
+    },
+    {
+      "epoch": 0.34362896017479394,
+      "grad_norm": 0.4072293871212264,
+      "learning_rate": 7.63618257043372e-06,
+      "loss": 0.0135,
+      "step": 1730
+    },
+    {
+      "epoch": 0.34382758963154236,
+      "grad_norm": 0.48094267145550845,
+      "learning_rate": 7.633448032958994e-06,
+      "loss": 0.0169,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3440262190882908,
+      "grad_norm": 0.546937299007096,
+      "learning_rate": 7.630712404976075e-06,
+      "loss": 0.0113,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3442248485450392,
+      "grad_norm": 0.595885300176229,
+      "learning_rate": 7.6279756876177835e-06,
+      "loss": 0.0156,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3444234780017877,
+      "grad_norm": 0.3227233399899523,
+      "learning_rate": 7.6252378820173915e-06,
+      "loss": 0.0064,
+      "step": 1734
+    },
+    {
+      "epoch": 0.3446221074585361,
+      "grad_norm": 0.6990835307410963,
+      "learning_rate": 7.622498989308622e-06,
+      "loss": 0.0197,
+      "step": 1735
+    },
+    {
+      "epoch": 0.34482073691528453,
+      "grad_norm": 0.67866449998373,
+      "learning_rate": 7.619759010625647e-06,
+      "loss": 0.0113,
+      "step": 1736
+    },
+    {
+      "epoch": 0.34501936637203295,
+      "grad_norm": 0.7839634602784449,
+      "learning_rate": 7.617017947103089e-06,
+      "loss": 0.0128,
+      "step": 1737
+    },
+    {
+      "epoch": 0.34521799582878143,
+      "grad_norm": 0.2870079380203802,
+      "learning_rate": 7.614275799876021e-06,
+      "loss": 0.0045,
+      "step": 1738
+    },
+    {
+      "epoch": 0.34541662528552985,
+      "grad_norm": 0.4689823234820678,
+      "learning_rate": 7.61153257007996e-06,
+      "loss": 0.0104,
+      "step": 1739
+    },
+    {
+      "epoch": 0.3456152547422783,
+      "grad_norm": 0.4687644833338894,
+      "learning_rate": 7.608788258850879e-06,
+      "loss": 0.0116,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3458138841990267,
+      "grad_norm": 0.5581341559370254,
+      "learning_rate": 7.6060428673251915e-06,
+      "loss": 0.0102,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3460125136557752,
+      "grad_norm": 0.5745297067207734,
+      "learning_rate": 7.603296396639763e-06,
+      "loss": 0.0129,
+      "step": 1742
+    },
+    {
+      "epoch": 0.3462111431125236,
+      "grad_norm": 0.5785097862293327,
+      "learning_rate": 7.600548847931903e-06,
+      "loss": 0.0101,
+      "step": 1743
+    },
+    {
+      "epoch": 0.346409772569272,
+      "grad_norm": 0.6057076832337547,
+      "learning_rate": 7.597800222339371e-06,
+      "loss": 0.0113,
+      "step": 1744
+    },
+    {
+      "epoch": 0.34660840202602045,
+      "grad_norm": 0.7114365255059848,
+      "learning_rate": 7.595050521000367e-06,
+      "loss": 0.0147,
+      "step": 1745
+    },
+    {
+      "epoch": 0.34680703148276887,
+      "grad_norm": 0.9549334149084857,
+      "learning_rate": 7.5922997450535405e-06,
+      "loss": 0.0079,
+      "step": 1746
+    },
+    {
+      "epoch": 0.34700566093951735,
+      "grad_norm": 0.9194225699846286,
+      "learning_rate": 7.589547895637987e-06,
+      "loss": 0.0176,
+      "step": 1747
+    },
+    {
+      "epoch": 0.34720429039626577,
+      "grad_norm": 0.317751182542115,
+      "learning_rate": 7.586794973893241e-06,
+      "loss": 0.0062,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3474029198530142,
+      "grad_norm": 0.7272366693065906,
+      "learning_rate": 7.584040980959288e-06,
+      "loss": 0.0114,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3476015493097626,
+      "grad_norm": 0.5400337381754199,
+      "learning_rate": 7.5812859179765555e-06,
+      "loss": 0.0084,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3478001787665111,
+      "grad_norm": 1.0261218838104105,
+      "learning_rate": 7.578529786085904e-06,
+      "loss": 0.0144,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3479988082232595,
+      "grad_norm": 0.5694769536171905,
+      "learning_rate": 7.5757725864286536e-06,
+      "loss": 0.0104,
+      "step": 1752
+    },
+    {
+      "epoch": 0.34819743768000794,
+      "grad_norm": 0.521723957173526,
+      "learning_rate": 7.573014320146554e-06,
+      "loss": 0.0119,
+      "step": 1753
+    },
+    {
+      "epoch": 0.34839606713675636,
+      "grad_norm": 0.6919236767327321,
+      "learning_rate": 7.570254988381801e-06,
+      "loss": 0.0134,
+      "step": 1754
+    },
+    {
+      "epoch": 0.34859469659350484,
+      "grad_norm": 0.5814649266815375,
+      "learning_rate": 7.567494592277031e-06,
+      "loss": 0.017,
+      "step": 1755
+    },
+    {
+      "epoch": 0.34879332605025326,
+      "grad_norm": 0.5249271812496219,
+      "learning_rate": 7.564733132975321e-06,
+      "loss": 0.011,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3489919555070017,
+      "grad_norm": 0.3963080625064441,
+      "learning_rate": 7.561970611620191e-06,
+      "loss": 0.0111,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3491905849637501,
+      "grad_norm": 0.7223775018258682,
+      "learning_rate": 7.559207029355593e-06,
+      "loss": 0.0173,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3493892144204986,
+      "grad_norm": 0.50134557033894,
+      "learning_rate": 7.5564423873259306e-06,
+      "loss": 0.0104,
+      "step": 1759
+    },
+    {
+      "epoch": 0.349587843877247,
+      "grad_norm": 0.6488452488216098,
+      "learning_rate": 7.553676686676034e-06,
+      "loss": 0.0175,
+      "step": 1760
+    },
+    {
+      "epoch": 0.34978647333399543,
+      "grad_norm": 0.49384847104699575,
+      "learning_rate": 7.5509099285511775e-06,
+      "loss": 0.0141,
+      "step": 1761
+    },
+    {
+      "epoch": 0.34998510279074385,
+      "grad_norm": 0.3953416148883208,
+      "learning_rate": 7.548142114097077e-06,
+      "loss": 0.012,
+      "step": 1762
+    },
+    {
+      "epoch": 0.35018373224749233,
+      "grad_norm": 0.666969527433996,
+      "learning_rate": 7.545373244459877e-06,
+      "loss": 0.0129,
+      "step": 1763
+    },
+    {
+      "epoch": 0.35038236170424075,
+      "grad_norm": 0.561354252191077,
+      "learning_rate": 7.542603320786166e-06,
+      "loss": 0.0208,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3505809911609892,
+      "grad_norm": 0.6255119965679132,
+      "learning_rate": 7.539832344222966e-06,
+      "loss": 0.0152,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3507796206177376,
+      "grad_norm": 0.4306557341184174,
+      "learning_rate": 7.537060315917734e-06,
+      "loss": 0.0211,
+      "step": 1766
+    },
+    {
+      "epoch": 0.350978250074486,
+      "grad_norm": 0.5528048136181183,
+      "learning_rate": 7.53428723701837e-06,
+      "loss": 0.0238,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3511768795312345,
+      "grad_norm": 0.5116386343446967,
+      "learning_rate": 7.531513108673196e-06,
+      "loss": 0.0172,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3513755089879829,
+      "grad_norm": 0.4279479662462199,
+      "learning_rate": 7.528737932030978e-06,
+      "loss": 0.0134,
+      "step": 1769
+    },
+    {
+      "epoch": 0.35157413844473134,
+      "grad_norm": 0.3728795654509069,
+      "learning_rate": 7.5259617082409165e-06,
+      "loss": 0.016,
+      "step": 1770
+    },
+    {
+      "epoch": 0.35177276790147977,
+      "grad_norm": 0.3025444908515605,
+      "learning_rate": 7.52318443845264e-06,
+      "loss": 0.0122,
+      "step": 1771
+    },
+    {
+      "epoch": 0.35197139735822824,
+      "grad_norm": 0.5225986004045365,
+      "learning_rate": 7.520406123816215e-06,
+      "loss": 0.0126,
+      "step": 1772
+    },
+    {
+      "epoch": 0.35217002681497667,
+      "grad_norm": 0.22475056903545595,
+      "learning_rate": 7.517626765482139e-06,
+      "loss": 0.0075,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3523686562717251,
+      "grad_norm": 0.6644702771927852,
+      "learning_rate": 7.5148463646013405e-06,
+      "loss": 0.0201,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3525672857284735,
+      "grad_norm": 0.21757342206133162,
+      "learning_rate": 7.512064922325179e-06,
+      "loss": 0.007,
+      "step": 1775
+    },
+    {
+      "epoch": 0.352765915185222,
+      "grad_norm": 0.7067021760903283,
+      "learning_rate": 7.50928243980545e-06,
+      "loss": 0.014,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3529645446419704,
+      "grad_norm": 0.6595482172027973,
+      "learning_rate": 7.506498918194376e-06,
+      "loss": 0.0117,
+      "step": 1777
+    },
+    {
+      "epoch": 0.35316317409871884,
+      "grad_norm": 0.6788932987088132,
+      "learning_rate": 7.5037143586446095e-06,
+      "loss": 0.0179,
+      "step": 1778
+    },
+    {
+      "epoch": 0.35336180355546726,
+      "grad_norm": 0.28064297644399,
+      "learning_rate": 7.500928762309234e-06,
+      "loss": 0.0086,
+      "step": 1779
+    },
+    {
+      "epoch": 0.35356043301221574,
+      "grad_norm": 0.5596372651009542,
+      "learning_rate": 7.498142130341764e-06,
+      "loss": 0.0166,
+      "step": 1780
+    },
+    {
+      "epoch": 0.35375906246896416,
+      "grad_norm": 0.7073177849186396,
+      "learning_rate": 7.495354463896137e-06,
+      "loss": 0.0095,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3539576919257126,
+      "grad_norm": 0.5066305988153633,
+      "learning_rate": 7.492565764126728e-06,
+      "loss": 0.0104,
+      "step": 1782
+    },
+    {
+      "epoch": 0.354156321382461,
+      "grad_norm": 0.44054068142958164,
+      "learning_rate": 7.4897760321883295e-06,
+      "loss": 0.0085,
+      "step": 1783
+    },
+    {
+      "epoch": 0.3543549508392095,
+      "grad_norm": 0.4553697052742875,
+      "learning_rate": 7.486985269236171e-06,
+      "loss": 0.0113,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3545535802959579,
+      "grad_norm": 0.5212306828856347,
+      "learning_rate": 7.4841934764259025e-06,
+      "loss": 0.0182,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3547522097527063,
+      "grad_norm": 0.31537340873334424,
+      "learning_rate": 7.481400654913606e-06,
+      "loss": 0.0095,
+      "step": 1786
+    },
+    {
+      "epoch": 0.35495083920945475,
+      "grad_norm": 0.7414978066014976,
+      "learning_rate": 7.47860680585578e-06,
+      "loss": 0.0207,
+      "step": 1787
+    },
+    {
+      "epoch": 0.35514946866620317,
+      "grad_norm": 0.6019848486242421,
+      "learning_rate": 7.475811930409359e-06,
+      "loss": 0.0157,
+      "step": 1788
+    },
+    {
+      "epoch": 0.35534809812295165,
+      "grad_norm": 0.3030605279151099,
+      "learning_rate": 7.473016029731696e-06,
+      "loss": 0.0067,
+      "step": 1789
+    },
+    {
+      "epoch": 0.3555467275797001,
+      "grad_norm": 1.415851763954458,
+      "learning_rate": 7.470219104980572e-06,
+      "loss": 0.0157,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3557453570364485,
+      "grad_norm": 0.3333734372525097,
+      "learning_rate": 7.467421157314191e-06,
+      "loss": 0.0102,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3559439864931969,
+      "grad_norm": 0.3295733399959379,
+      "learning_rate": 7.464622187891179e-06,
+      "loss": 0.0093,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3561426159499454,
+      "grad_norm": 0.5348317682346659,
+      "learning_rate": 7.4618221978705875e-06,
+      "loss": 0.0225,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3563412454066938,
+      "grad_norm": 0.4680465985095257,
+      "learning_rate": 7.45902118841189e-06,
+      "loss": 0.0098,
+      "step": 1794
+    },
+    {
+      "epoch": 0.35653987486344224,
+      "grad_norm": 0.5680929119198581,
+      "learning_rate": 7.45621916067498e-06,
+      "loss": 0.0154,
+      "step": 1795
+    },
+    {
+      "epoch": 0.35673850432019066,
+      "grad_norm": 0.7557185174597132,
+      "learning_rate": 7.453416115820173e-06,
+      "loss": 0.0157,
+      "step": 1796
+    },
+    {
+      "epoch": 0.35693713377693914,
+      "grad_norm": 0.6641326101462152,
+      "learning_rate": 7.4506120550082125e-06,
+      "loss": 0.0158,
+      "step": 1797
+    },
+    {
+      "epoch": 0.35713576323368756,
+      "grad_norm": 0.7831104663272818,
+      "learning_rate": 7.447806979400255e-06,
+      "loss": 0.0126,
+      "step": 1798
+    },
+    {
+      "epoch": 0.357334392690436,
+      "grad_norm": 0.4053205882012716,
+      "learning_rate": 7.445000890157876e-06,
+      "loss": 0.0114,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3575330221471844,
+      "grad_norm": 0.5298271130207239,
+      "learning_rate": 7.442193788443078e-06,
+      "loss": 0.0094,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3577316516039329,
+      "grad_norm": 0.9078634591414197,
+      "learning_rate": 7.439385675418278e-06,
+      "loss": 0.0128,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3579302810606813,
+      "grad_norm": 0.4697586290419777,
+      "learning_rate": 7.436576552246312e-06,
+      "loss": 0.0103,
+      "step": 1802
+    },
+    {
+      "epoch": 0.35812891051742973,
+      "grad_norm": 0.9172712379027379,
+      "learning_rate": 7.433766420090436e-06,
+      "loss": 0.0198,
+      "step": 1803
+    },
+    {
+      "epoch": 0.35832753997417816,
+      "grad_norm": 0.45567523468910553,
+      "learning_rate": 7.430955280114322e-06,
+      "loss": 0.0119,
+      "step": 1804
+    },
+    {
+      "epoch": 0.35852616943092663,
+      "grad_norm": 0.7802133005429396,
+      "learning_rate": 7.428143133482063e-06,
+      "loss": 0.009,
+      "step": 1805
+    },
+    {
+      "epoch": 0.35872479888767506,
+      "grad_norm": 0.6423209564757857,
+      "learning_rate": 7.425329981358163e-06,
+      "loss": 0.0214,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3589234283444235,
+      "grad_norm": 0.49872407430644544,
+      "learning_rate": 7.422515824907546e-06,
+      "loss": 0.0155,
+      "step": 1807
+    },
+    {
+      "epoch": 0.3591220578011719,
+      "grad_norm": 0.40918220122528187,
+      "learning_rate": 7.419700665295551e-06,
+      "loss": 0.0113,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3593206872579203,
+      "grad_norm": 0.4717139361458346,
+      "learning_rate": 7.416884503687936e-06,
+      "loss": 0.0089,
+      "step": 1809
+    },
+    {
+      "epoch": 0.3595193167146688,
+      "grad_norm": 0.7744599282851102,
+      "learning_rate": 7.414067341250868e-06,
+      "loss": 0.0175,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3597179461714172,
+      "grad_norm": 0.7426285798479693,
+      "learning_rate": 7.41124917915093e-06,
+      "loss": 0.018,
+      "step": 1811
+    },
+    {
+      "epoch": 0.35991657562816565,
+      "grad_norm": 0.3054174866414779,
+      "learning_rate": 7.408430018555122e-06,
+      "loss": 0.0108,
+      "step": 1812
+    },
+    {
+      "epoch": 0.36011520508491407,
+      "grad_norm": 0.44806274127264745,
+      "learning_rate": 7.405609860630855e-06,
+      "loss": 0.0159,
+      "step": 1813
+    },
+    {
+      "epoch": 0.36031383454166255,
+      "grad_norm": 0.45456734941331584,
+      "learning_rate": 7.402788706545953e-06,
+      "loss": 0.0114,
+      "step": 1814
+    },
+    {
+      "epoch": 0.36051246399841097,
+      "grad_norm": 0.41653981798920153,
+      "learning_rate": 7.3999665574686566e-06,
+      "loss": 0.0167,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3607110934551594,
+      "grad_norm": 0.42591262940535,
+      "learning_rate": 7.39714341456761e-06,
+      "loss": 0.012,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3609097229119078,
+      "grad_norm": 0.5559851848169303,
+      "learning_rate": 7.394319279011877e-06,
+      "loss": 0.016,
+      "step": 1817
+    },
+    {
+      "epoch": 0.3611083523686563,
+      "grad_norm": 0.3959904663844897,
+      "learning_rate": 7.391494151970928e-06,
+      "loss": 0.0123,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3613069818254047,
+      "grad_norm": 0.6179966278625685,
+      "learning_rate": 7.388668034614645e-06,
+      "loss": 0.0135,
+      "step": 1819
+    },
+    {
+      "epoch": 0.36150561128215314,
+      "grad_norm": 0.37571678679572235,
+      "learning_rate": 7.385840928113321e-06,
+      "loss": 0.01,
+      "step": 1820
+    },
+    {
+      "epoch": 0.36170424073890156,
+      "grad_norm": 0.4874416376117431,
+      "learning_rate": 7.383012833637657e-06,
+      "loss": 0.0111,
+      "step": 1821
+    },
+    {
+      "epoch": 0.36190287019565004,
+      "grad_norm": 0.4693254096635751,
+      "learning_rate": 7.380183752358768e-06,
+      "loss": 0.0154,
+      "step": 1822
+    },
+    {
+      "epoch": 0.36210149965239846,
+      "grad_norm": 0.6119621192596436,
+      "learning_rate": 7.37735368544817e-06,
+      "loss": 0.0187,
+      "step": 1823
+    },
+    {
+      "epoch": 0.3623001291091469,
+      "grad_norm": 0.35374113186809264,
+      "learning_rate": 7.37452263407779e-06,
+      "loss": 0.0102,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3624987585658953,
+      "grad_norm": 0.7498478793443032,
+      "learning_rate": 7.371690599419965e-06,
+      "loss": 0.0147,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3626973880226438,
+      "grad_norm": 0.9295665798522207,
+      "learning_rate": 7.3688575826474385e-06,
+      "loss": 0.026,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3628960174793922,
+      "grad_norm": 0.523390784697274,
+      "learning_rate": 7.3660235849333594e-06,
+      "loss": 0.0169,
+      "step": 1827
+    },
+    {
+      "epoch": 0.36309464693614063,
+      "grad_norm": 0.4162831159506049,
+      "learning_rate": 7.363188607451283e-06,
+      "loss": 0.0132,
+      "step": 1828
+    },
+    {
+      "epoch": 0.36329327639288905,
+      "grad_norm": 0.6182223304228157,
+      "learning_rate": 7.360352651375171e-06,
+      "loss": 0.0138,
+      "step": 1829
+    },
+    {
+      "epoch": 0.3634919058496375,
+      "grad_norm": 0.43001898999202587,
+      "learning_rate": 7.35751571787939e-06,
+      "loss": 0.0137,
+      "step": 1830
+    },
+    {
+      "epoch": 0.36369053530638595,
+      "grad_norm": 0.5552465012209282,
+      "learning_rate": 7.35467780813871e-06,
+      "loss": 0.0103,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3638891647631344,
+      "grad_norm": 0.5016437757951848,
+      "learning_rate": 7.3518389233283095e-06,
+      "loss": 0.0094,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3640877942198828,
+      "grad_norm": 0.18076930934715546,
+      "learning_rate": 7.348999064623763e-06,
+      "loss": 0.0059,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3642864236766312,
+      "grad_norm": 0.817939106262419,
+      "learning_rate": 7.34615823320106e-06,
+      "loss": 0.021,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3644850531333797,
+      "grad_norm": 0.517907031178373,
+      "learning_rate": 7.34331643023658e-06,
+      "loss": 0.0129,
+      "step": 1835
+    },
+    {
+      "epoch": 0.3646836825901281,
+      "grad_norm": 0.4513320397295181,
+      "learning_rate": 7.340473656907113e-06,
+      "loss": 0.0127,
+      "step": 1836
+    },
+    {
+      "epoch": 0.36488231204687654,
+      "grad_norm": 0.7014290051670106,
+      "learning_rate": 7.33762991438985e-06,
+      "loss": 0.0126,
+      "step": 1837
+    },
+    {
+      "epoch": 0.36508094150362497,
+      "grad_norm": 0.34486163648433743,
+      "learning_rate": 7.334785203862378e-06,
+      "loss": 0.0131,
+      "step": 1838
+    },
+    {
+      "epoch": 0.36527957096037345,
+      "grad_norm": 0.42658500377007785,
+      "learning_rate": 7.331939526502692e-06,
+      "loss": 0.0148,
+      "step": 1839
+    },
+    {
+      "epoch": 0.36547820041712187,
+      "grad_norm": 0.4012284928050507,
+      "learning_rate": 7.329092883489184e-06,
+      "loss": 0.0103,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3656768298738703,
+      "grad_norm": 0.6923841780226727,
+      "learning_rate": 7.326245276000645e-06,
+      "loss": 0.0173,
+      "step": 1841
+    },
+    {
+      "epoch": 0.3658754593306187,
+      "grad_norm": 0.6204656956557247,
+      "learning_rate": 7.323396705216267e-06,
+      "loss": 0.0149,
+      "step": 1842
+    },
+    {
+      "epoch": 0.3660740887873672,
+      "grad_norm": 0.43915467134030417,
+      "learning_rate": 7.320547172315639e-06,
+      "loss": 0.0115,
+      "step": 1843
+    },
+    {
+      "epoch": 0.3662727182441156,
+      "grad_norm": 0.9453561263857528,
+      "learning_rate": 7.317696678478752e-06,
+      "loss": 0.0147,
+      "step": 1844
+    },
+    {
+      "epoch": 0.36647134770086404,
+      "grad_norm": 0.4116346164383005,
+      "learning_rate": 7.314845224885992e-06,
+      "loss": 0.0132,
+      "step": 1845
+    },
+    {
+      "epoch": 0.36666997715761246,
+      "grad_norm": 0.6010586200072675,
+      "learning_rate": 7.31199281271814e-06,
+      "loss": 0.016,
+      "step": 1846
+    },
+    {
+      "epoch": 0.36686860661436094,
+      "grad_norm": 0.379756980977871,
+      "learning_rate": 7.309139443156382e-06,
+      "loss": 0.0084,
+      "step": 1847
+    },
+    {
+      "epoch": 0.36706723607110936,
+      "grad_norm": 0.32996713284661944,
+      "learning_rate": 7.306285117382292e-06,
+      "loss": 0.0078,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3672658655278578,
+      "grad_norm": 0.3528111052658405,
+      "learning_rate": 7.3034298365778455e-06,
+      "loss": 0.0105,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3674644949846062,
+      "grad_norm": 0.32743299157906436,
+      "learning_rate": 7.300573601925409e-06,
+      "loss": 0.0129,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3676631244413546,
+      "grad_norm": 0.40089814984231775,
+      "learning_rate": 7.297716414607747e-06,
+      "loss": 0.0074,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3678617538981031,
+      "grad_norm": 0.8879963171017308,
+      "learning_rate": 7.294858275808021e-06,
+      "loss": 0.0158,
+      "step": 1852
+    },
+    {
+      "epoch": 0.36806038335485153,
+      "grad_norm": 1.1591210573488786,
+      "learning_rate": 7.29199918670978e-06,
+      "loss": 0.0179,
+      "step": 1853
+    },
+    {
+      "epoch": 0.36825901281159995,
+      "grad_norm": 0.5766334292979398,
+      "learning_rate": 7.289139148496971e-06,
+      "loss": 0.0134,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3684576422683484,
+      "grad_norm": 0.5074699980363011,
+      "learning_rate": 7.286278162353934e-06,
+      "loss": 0.0087,
+      "step": 1855
+    },
+    {
+      "epoch": 0.36865627172509685,
+      "grad_norm": 0.5200342371854682,
+      "learning_rate": 7.283416229465399e-06,
+      "loss": 0.0154,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3688549011818453,
+      "grad_norm": 0.4688805862662405,
+      "learning_rate": 7.280553351016489e-06,
+      "loss": 0.0083,
+      "step": 1857
+    },
+    {
+      "epoch": 0.3690535306385937,
+      "grad_norm": 0.5049930399031888,
+      "learning_rate": 7.277689528192722e-06,
+      "loss": 0.0113,
+      "step": 1858
+    },
+    {
+      "epoch": 0.3692521600953421,
+      "grad_norm": 0.8576327838457353,
+      "learning_rate": 7.2748247621800005e-06,
+      "loss": 0.0151,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3694507895520906,
+      "grad_norm": 0.5565380252033648,
+      "learning_rate": 7.271959054164623e-06,
+      "loss": 0.0117,
+      "step": 1860
+    },
+    {
+      "epoch": 0.369649419008839,
+      "grad_norm": 1.193299984883163,
+      "learning_rate": 7.269092405333278e-06,
+      "loss": 0.0153,
+      "step": 1861
+    },
+    {
+      "epoch": 0.36984804846558744,
+      "grad_norm": 0.8109235962868375,
+      "learning_rate": 7.26622481687304e-06,
+      "loss": 0.0166,
+      "step": 1862
+    },
+    {
+      "epoch": 0.37004667792233586,
+      "grad_norm": 0.7632333804661434,
+      "learning_rate": 7.263356289971374e-06,
+      "loss": 0.0169,
+      "step": 1863
+    },
+    {
+      "epoch": 0.37024530737908434,
+      "grad_norm": 0.8232197885172541,
+      "learning_rate": 7.260486825816134e-06,
+      "loss": 0.0165,
+      "step": 1864
+    },
+    {
+      "epoch": 0.37044393683583277,
+      "grad_norm": 0.4215445953865816,
+      "learning_rate": 7.257616425595564e-06,
+      "loss": 0.0146,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3706425662925812,
+      "grad_norm": 0.6077339267039444,
+      "learning_rate": 7.254745090498294e-06,
+      "loss": 0.0179,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3708411957493296,
+      "grad_norm": 0.8156039287786392,
+      "learning_rate": 7.251872821713339e-06,
+      "loss": 0.015,
+      "step": 1867
+    },
+    {
+      "epoch": 0.3710398252060781,
+      "grad_norm": 0.4884625901216402,
+      "learning_rate": 7.248999620430104e-06,
+      "loss": 0.0076,
+      "step": 1868
+    },
+    {
+      "epoch": 0.3712384546628265,
+      "grad_norm": 0.47130775104617345,
+      "learning_rate": 7.246125487838378e-06,
+      "loss": 0.012,
+      "step": 1869
+    },
+    {
+      "epoch": 0.37143708411957493,
+      "grad_norm": 0.3343798674016324,
+      "learning_rate": 7.243250425128337e-06,
+      "loss": 0.0082,
+      "step": 1870
+    },
+    {
+      "epoch": 0.37163571357632336,
+      "grad_norm": 0.2962776732176791,
+      "learning_rate": 7.240374433490542e-06,
+      "loss": 0.0092,
+      "step": 1871
+    },
+    {
+      "epoch": 0.3718343430330718,
+      "grad_norm": 0.6443601303540771,
+      "learning_rate": 7.237497514115937e-06,
+      "loss": 0.0149,
+      "step": 1872
+    },
+    {
+      "epoch": 0.37203297248982026,
+      "grad_norm": 0.6090312171198041,
+      "learning_rate": 7.234619668195853e-06,
+      "loss": 0.022,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3722316019465687,
+      "grad_norm": 1.008498309205369,
+      "learning_rate": 7.231740896922e-06,
+      "loss": 0.0181,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3724302314033171,
+      "grad_norm": 0.2263113111319505,
+      "learning_rate": 7.228861201486479e-06,
+      "loss": 0.0107,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3726288608600655,
+      "grad_norm": 1.0824047330601112,
+      "learning_rate": 7.225980583081764e-06,
+      "loss": 0.015,
+      "step": 1876
+    },
+    {
+      "epoch": 0.372827490316814,
+      "grad_norm": 0.4519637815818529,
+      "learning_rate": 7.2230990429007205e-06,
+      "loss": 0.0167,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3730261197735624,
+      "grad_norm": 0.3996641096482069,
+      "learning_rate": 7.2202165821365884e-06,
+      "loss": 0.0099,
+      "step": 1878
+    },
+    {
+      "epoch": 0.37322474923031085,
+      "grad_norm": 0.5247153139606248,
+      "learning_rate": 7.217333201982994e-06,
+      "loss": 0.0112,
+      "step": 1879
+    },
+    {
+      "epoch": 0.37342337868705927,
+      "grad_norm": 0.6322765470012125,
+      "learning_rate": 7.2144489036339414e-06,
+      "loss": 0.01,
+      "step": 1880
+    },
+    {
+      "epoch": 0.37362200814380775,
+      "grad_norm": 0.7784772761848201,
+      "learning_rate": 7.211563688283815e-06,
+      "loss": 0.0093,
+      "step": 1881
+    },
+    {
+      "epoch": 0.37382063760055617,
+      "grad_norm": 0.489780522096192,
+      "learning_rate": 7.20867755712738e-06,
+      "loss": 0.0107,
+      "step": 1882
+    },
+    {
+      "epoch": 0.3740192670573046,
+      "grad_norm": 0.4222831983087626,
+      "learning_rate": 7.20579051135978e-06,
+      "loss": 0.0133,
+      "step": 1883
+    },
+    {
+      "epoch": 0.374217896514053,
+      "grad_norm": 0.39180257048360806,
+      "learning_rate": 7.2029025521765395e-06,
+      "loss": 0.0084,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3744165259708015,
+      "grad_norm": 0.328589453430018,
+      "learning_rate": 7.200013680773556e-06,
+      "loss": 0.0086,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3746151554275499,
+      "grad_norm": 0.8840476580346277,
+      "learning_rate": 7.197123898347113e-06,
+      "loss": 0.0197,
+      "step": 1886
+    },
+    {
+      "epoch": 0.37481378488429834,
+      "grad_norm": 0.7445076848470832,
+      "learning_rate": 7.194233206093862e-06,
+      "loss": 0.0155,
+      "step": 1887
+    },
+    {
+      "epoch": 0.37501241434104676,
+      "grad_norm": 0.823712336571771,
+      "learning_rate": 7.1913416052108385e-06,
+      "loss": 0.0138,
+      "step": 1888
+    },
+    {
+      "epoch": 0.37521104379779524,
+      "grad_norm": 0.5026615807876057,
+      "learning_rate": 7.18844909689545e-06,
+      "loss": 0.0127,
+      "step": 1889
+    },
+    {
+      "epoch": 0.37540967325454366,
+      "grad_norm": 0.5235102842153762,
+      "learning_rate": 7.185555682345483e-06,
+      "loss": 0.0167,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3756083027112921,
+      "grad_norm": 0.38122765851314344,
+      "learning_rate": 7.182661362759096e-06,
+      "loss": 0.016,
+      "step": 1891
+    },
+    {
+      "epoch": 0.3758069321680405,
+      "grad_norm": 0.659952705620418,
+      "learning_rate": 7.179766139334825e-06,
+      "loss": 0.0217,
+      "step": 1892
+    },
+    {
+      "epoch": 0.37600556162478893,
+      "grad_norm": 0.3460455608781431,
+      "learning_rate": 7.1768700132715785e-06,
+      "loss": 0.0122,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3762041910815374,
+      "grad_norm": 0.5470309221958665,
+      "learning_rate": 7.173972985768639e-06,
+      "loss": 0.0159,
+      "step": 1894
+    },
+    {
+      "epoch": 0.37640282053828583,
+      "grad_norm": 0.609575691356759,
+      "learning_rate": 7.171075058025664e-06,
+      "loss": 0.015,
+      "step": 1895
+    },
+    {
+      "epoch": 0.37660144999503425,
+      "grad_norm": 0.6256172821754186,
+      "learning_rate": 7.168176231242681e-06,
+      "loss": 0.0091,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3768000794517827,
+      "grad_norm": 0.43669928645159256,
+      "learning_rate": 7.165276506620092e-06,
+      "loss": 0.0108,
+      "step": 1897
+    },
+    {
+      "epoch": 0.37699870890853115,
+      "grad_norm": 0.452629904550346,
+      "learning_rate": 7.16237588535867e-06,
+      "loss": 0.0131,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3771973383652796,
+      "grad_norm": 0.5306623947422398,
+      "learning_rate": 7.159474368659559e-06,
+      "loss": 0.0101,
+      "step": 1899
+    },
+    {
+      "epoch": 0.377395967822028,
+      "grad_norm": 0.877007063554911,
+      "learning_rate": 7.156571957724275e-06,
+      "loss": 0.0221,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3775945972787764,
+      "grad_norm": 0.5158241756842965,
+      "learning_rate": 7.153668653754702e-06,
+      "loss": 0.0129,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3777932267355249,
+      "grad_norm": 0.3126658544481895,
+      "learning_rate": 7.150764457953096e-06,
+      "loss": 0.0131,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3779918561922733,
+      "grad_norm": 0.8119239874875152,
+      "learning_rate": 7.147859371522083e-06,
+      "loss": 0.0149,
+      "step": 1903
+    },
+    {
+      "epoch": 0.37819048564902175,
+      "grad_norm": 0.37864224886279996,
+      "learning_rate": 7.1449533956646555e-06,
+      "loss": 0.0166,
+      "step": 1904
+    },
+    {
+      "epoch": 0.37838911510577017,
+      "grad_norm": 0.31041396376686636,
+      "learning_rate": 7.142046531584176e-06,
+      "loss": 0.0077,
+      "step": 1905
+    },
+    {
+      "epoch": 0.37858774456251865,
+      "grad_norm": 0.44678399616930514,
+      "learning_rate": 7.139138780484371e-06,
+      "loss": 0.0147,
+      "step": 1906
+    },
+    {
+      "epoch": 0.37878637401926707,
+      "grad_norm": 0.6814421963342644,
+      "learning_rate": 7.136230143569338e-06,
+      "loss": 0.0092,
+      "step": 1907
+    },
+    {
+      "epoch": 0.3789850034760155,
+      "grad_norm": 0.5699394737446986,
+      "learning_rate": 7.133320622043544e-06,
+      "loss": 0.0095,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3791836329327639,
+      "grad_norm": 0.5118628899874634,
+      "learning_rate": 7.1304102171118165e-06,
+      "loss": 0.018,
+      "step": 1909
+    },
+    {
+      "epoch": 0.37938226238951234,
+      "grad_norm": 0.8450993817197867,
+      "learning_rate": 7.12749892997935e-06,
+      "loss": 0.0152,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3795808918462608,
+      "grad_norm": 0.5936810791362285,
+      "learning_rate": 7.124586761851709e-06,
+      "loss": 0.0137,
+      "step": 1911
+    },
+    {
+      "epoch": 0.37977952130300924,
+      "grad_norm": 0.7147959398627135,
+      "learning_rate": 7.121673713934816e-06,
+      "loss": 0.0075,
+      "step": 1912
+    },
+    {
+      "epoch": 0.37997815075975766,
+      "grad_norm": 0.3268090768205218,
+      "learning_rate": 7.1187597874349635e-06,
+      "loss": 0.0086,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3801767802165061,
+      "grad_norm": 0.8429071661543824,
+      "learning_rate": 7.115844983558804e-06,
+      "loss": 0.0164,
+      "step": 1914
+    },
+    {
+      "epoch": 0.38037540967325456,
+      "grad_norm": 0.5508174982198967,
+      "learning_rate": 7.112929303513356e-06,
+      "loss": 0.0185,
+      "step": 1915
+    },
+    {
+      "epoch": 0.380574039130003,
+      "grad_norm": 0.8107547717459483,
+      "learning_rate": 7.110012748506e-06,
+      "loss": 0.0191,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3807726685867514,
+      "grad_norm": 0.4591386080076129,
+      "learning_rate": 7.107095319744479e-06,
+      "loss": 0.0108,
+      "step": 1917
+    },
+    {
+      "epoch": 0.38097129804349983,
+      "grad_norm": 0.7765034782212996,
+      "learning_rate": 7.1041770184368945e-06,
+      "loss": 0.0119,
+      "step": 1918
+    },
+    {
+      "epoch": 0.3811699275002483,
+      "grad_norm": 0.5274182540525633,
+      "learning_rate": 7.101257845791714e-06,
+      "loss": 0.0141,
+      "step": 1919
+    },
+    {
+      "epoch": 0.38136855695699673,
+      "grad_norm": 0.3603680268183808,
+      "learning_rate": 7.098337803017763e-06,
+      "loss": 0.0087,
+      "step": 1920
+    },
+    {
+      "epoch": 0.38156718641374515,
+      "grad_norm": 0.47632384862279487,
+      "learning_rate": 7.095416891324231e-06,
+      "loss": 0.0083,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3817658158704936,
+      "grad_norm": 0.6379591206386355,
+      "learning_rate": 7.0924951119206605e-06,
+      "loss": 0.0113,
+      "step": 1922
+    },
+    {
+      "epoch": 0.38196444532724205,
+      "grad_norm": 0.5608777505358337,
+      "learning_rate": 7.0895724660169615e-06,
+      "loss": 0.0159,
+      "step": 1923
+    },
+    {
+      "epoch": 0.3821630747839905,
+      "grad_norm": 0.45075216436612026,
+      "learning_rate": 7.086648954823396e-06,
+      "loss": 0.0103,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3823617042407389,
+      "grad_norm": 0.8387316921379768,
+      "learning_rate": 7.083724579550588e-06,
+      "loss": 0.0248,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3825603336974873,
+      "grad_norm": 0.504954461339415,
+      "learning_rate": 7.080799341409518e-06,
+      "loss": 0.0162,
+      "step": 1926
+    },
+    {
+      "epoch": 0.3827589631542358,
+      "grad_norm": 0.633175612122431,
+      "learning_rate": 7.077873241611525e-06,
+      "loss": 0.0134,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3829575926109842,
+      "grad_norm": 0.5726947868855516,
+      "learning_rate": 7.074946281368304e-06,
+      "loss": 0.0154,
+      "step": 1928
+    },
+    {
+      "epoch": 0.38315622206773264,
+      "grad_norm": 0.27366180430338843,
+      "learning_rate": 7.072018461891906e-06,
+      "loss": 0.0122,
+      "step": 1929
+    },
+    {
+      "epoch": 0.38335485152448107,
+      "grad_norm": 0.3847532093947302,
+      "learning_rate": 7.069089784394737e-06,
+      "loss": 0.01,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3835534809812295,
+      "grad_norm": 0.40845059746079004,
+      "learning_rate": 7.066160250089561e-06,
+      "loss": 0.0101,
+      "step": 1931
+    },
+    {
+      "epoch": 0.38375211043797797,
+      "grad_norm": 0.4485746408020155,
+      "learning_rate": 7.063229860189493e-06,
+      "loss": 0.0105,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3839507398947264,
+      "grad_norm": 0.9400839412508927,
+      "learning_rate": 7.060298615908006e-06,
+      "loss": 0.0153,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3841493693514748,
+      "grad_norm": 0.3182534498151112,
+      "learning_rate": 7.057366518458928e-06,
+      "loss": 0.0111,
+      "step": 1934
+    },
+    {
+      "epoch": 0.38434799880822323,
+      "grad_norm": 0.4800246818941992,
+      "learning_rate": 7.0544335690564334e-06,
+      "loss": 0.0209,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3845466282649717,
+      "grad_norm": 0.46973294531377413,
+      "learning_rate": 7.051499768915056e-06,
+      "loss": 0.0114,
+      "step": 1936
+    },
+    {
+      "epoch": 0.38474525772172014,
+      "grad_norm": 0.7859855280724575,
+      "learning_rate": 7.048565119249677e-06,
+      "loss": 0.0141,
+      "step": 1937
+    },
+    {
+      "epoch": 0.38494388717846856,
+      "grad_norm": 0.6876110737606786,
+      "learning_rate": 7.0456296212755344e-06,
+      "loss": 0.0128,
+      "step": 1938
+    },
+    {
+      "epoch": 0.385142516635217,
+      "grad_norm": 0.46899691256126286,
+      "learning_rate": 7.0426932762082135e-06,
+      "loss": 0.0093,
+      "step": 1939
+    },
+    {
+      "epoch": 0.38534114609196546,
+      "grad_norm": 0.3513359380168546,
+      "learning_rate": 7.039756085263654e-06,
+      "loss": 0.0072,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3855397755487139,
+      "grad_norm": 0.6075133338918507,
+      "learning_rate": 7.036818049658143e-06,
+      "loss": 0.0143,
+      "step": 1941
+    },
+    {
+      "epoch": 0.3857384050054623,
+      "grad_norm": 0.7662029488077239,
+      "learning_rate": 7.0338791706083155e-06,
+      "loss": 0.0132,
+      "step": 1942
+    },
+    {
+      "epoch": 0.3859370344622107,
+      "grad_norm": 0.628109956535636,
+      "learning_rate": 7.03093944933116e-06,
+      "loss": 0.016,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3861356639189592,
+      "grad_norm": 0.19803867516817825,
+      "learning_rate": 7.027998887044013e-06,
+      "loss": 0.0052,
+      "step": 1944
+    },
+    {
+      "epoch": 0.3863342933757076,
+      "grad_norm": 0.662751247765471,
+      "learning_rate": 7.025057484964558e-06,
+      "loss": 0.0104,
+      "step": 1945
+    },
+    {
+      "epoch": 0.38653292283245605,
+      "grad_norm": 0.6819086619122386,
+      "learning_rate": 7.022115244310826e-06,
+      "loss": 0.0113,
+      "step": 1946
+    },
+    {
+      "epoch": 0.38673155228920447,
+      "grad_norm": 1.5798621153907724,
+      "learning_rate": 7.019172166301197e-06,
+      "loss": 0.0223,
+      "step": 1947
+    },
+    {
+      "epoch": 0.38693018174595295,
+      "grad_norm": 0.5987945163370364,
+      "learning_rate": 7.016228252154396e-06,
+      "loss": 0.0121,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3871288112027014,
+      "grad_norm": 0.9981941212502284,
+      "learning_rate": 7.013283503089494e-06,
+      "loss": 0.0115,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3873274406594498,
+      "grad_norm": 1.0406382038091888,
+      "learning_rate": 7.01033792032591e-06,
+      "loss": 0.0248,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3875260701161982,
+      "grad_norm": 0.41775245079601303,
+      "learning_rate": 7.007391505083405e-06,
+      "loss": 0.0104,
+      "step": 1951
+    },
+    {
+      "epoch": 0.38772469957294664,
+      "grad_norm": 0.5898166538524512,
+      "learning_rate": 7.004444258582091e-06,
+      "loss": 0.0129,
+      "step": 1952
+    },
+    {
+      "epoch": 0.3879233290296951,
+      "grad_norm": 0.7611476458360803,
+      "learning_rate": 7.001496182042416e-06,
+      "loss": 0.0157,
+      "step": 1953
+    },
+    {
+      "epoch": 0.38812195848644354,
+      "grad_norm": 0.33055910901797114,
+      "learning_rate": 6.998547276685175e-06,
+      "loss": 0.0107,
+      "step": 1954
+    },
+    {
+      "epoch": 0.38832058794319196,
+      "grad_norm": 0.48354068516638127,
+      "learning_rate": 6.995597543731509e-06,
+      "loss": 0.0085,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3885192173999404,
+      "grad_norm": 1.3859038603039011,
+      "learning_rate": 6.992646984402898e-06,
+      "loss": 0.022,
+      "step": 1956
+    },
+    {
+      "epoch": 0.38871784685668886,
+      "grad_norm": 0.9177731485118693,
+      "learning_rate": 6.989695599921166e-06,
+      "loss": 0.0192,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3889164763134373,
+      "grad_norm": 0.6987553756434028,
+      "learning_rate": 6.986743391508479e-06,
+      "loss": 0.0182,
+      "step": 1958
+    },
+    {
+      "epoch": 0.3891151057701857,
+      "grad_norm": 0.40997833208509904,
+      "learning_rate": 6.983790360387344e-06,
+      "loss": 0.0088,
+      "step": 1959
+    },
+    {
+      "epoch": 0.38931373522693413,
+      "grad_norm": 0.34916159941815383,
+      "learning_rate": 6.980836507780606e-06,
+      "loss": 0.0128,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3895123646836826,
+      "grad_norm": 0.7820977108647957,
+      "learning_rate": 6.977881834911455e-06,
+      "loss": 0.0122,
+      "step": 1961
+    },
+    {
+      "epoch": 0.38971099414043103,
+      "grad_norm": 0.9305707403285578,
+      "learning_rate": 6.974926343003416e-06,
+      "loss": 0.0128,
+      "step": 1962
+    },
+    {
+      "epoch": 0.38990962359717946,
+      "grad_norm": 0.682174849117662,
+      "learning_rate": 6.971970033280356e-06,
+      "loss": 0.0108,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3901082530539279,
+      "grad_norm": 0.5350787998659173,
+      "learning_rate": 6.96901290696648e-06,
+      "loss": 0.0154,
+      "step": 1964
+    },
+    {
+      "epoch": 0.39030688251067636,
+      "grad_norm": 0.3819759718576897,
+      "learning_rate": 6.966054965286333e-06,
+      "loss": 0.0098,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3905055119674248,
+      "grad_norm": 0.5388430417097986,
+      "learning_rate": 6.963096209464793e-06,
+      "loss": 0.0144,
+      "step": 1966
+    },
+    {
+      "epoch": 0.3907041414241732,
+      "grad_norm": 0.5604706168748357,
+      "learning_rate": 6.9601366407270766e-06,
+      "loss": 0.0157,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3909027708809216,
+      "grad_norm": 0.5074402312211344,
+      "learning_rate": 6.957176260298742e-06,
+      "loss": 0.008,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3911014003376701,
+      "grad_norm": 0.37696620230600386,
+      "learning_rate": 6.954215069405677e-06,
+      "loss": 0.0164,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3913000297944185,
+      "grad_norm": 0.7845133541601853,
+      "learning_rate": 6.951253069274109e-06,
+      "loss": 0.013,
+      "step": 1970
+    },
+    {
+      "epoch": 0.39149865925116695,
+      "grad_norm": 0.6377358194064576,
+      "learning_rate": 6.948290261130598e-06,
+      "loss": 0.0108,
+      "step": 1971
+    },
+    {
+      "epoch": 0.39169728870791537,
+      "grad_norm": 0.37305516874735534,
+      "learning_rate": 6.945326646202042e-06,
+      "loss": 0.0118,
+      "step": 1972
+    },
+    {
+      "epoch": 0.3918959181646638,
+      "grad_norm": 0.6101758865406128,
+      "learning_rate": 6.942362225715671e-06,
+      "loss": 0.0072,
+      "step": 1973
+    },
+    {
+      "epoch": 0.39209454762141227,
+      "grad_norm": 0.6669153608848816,
+      "learning_rate": 6.939397000899046e-06,
+      "loss": 0.0164,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3922931770781607,
+      "grad_norm": 0.4434397317662605,
+      "learning_rate": 6.936430972980066e-06,
+      "loss": 0.0142,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3924918065349091,
+      "grad_norm": 1.1486290060634483,
+      "learning_rate": 6.933464143186958e-06,
+      "loss": 0.0246,
+      "step": 1976
+    },
+    {
+      "epoch": 0.39269043599165754,
+      "grad_norm": 0.4402208692427919,
+      "learning_rate": 6.930496512748286e-06,
+      "loss": 0.0142,
+      "step": 1977
+    },
+    {
+      "epoch": 0.392889065448406,
+      "grad_norm": 0.8605918213090531,
+      "learning_rate": 6.927528082892941e-06,
+      "loss": 0.0162,
+      "step": 1978
+    },
+    {
+      "epoch": 0.39308769490515444,
+      "grad_norm": 0.602808108046716,
+      "learning_rate": 6.924558854850146e-06,
+      "loss": 0.0202,
+      "step": 1979
+    },
+    {
+      "epoch": 0.39328632436190286,
+      "grad_norm": 1.044860369928853,
+      "learning_rate": 6.921588829849458e-06,
+      "loss": 0.0239,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3934849538186513,
+      "grad_norm": 1.0031230137699616,
+      "learning_rate": 6.918618009120759e-06,
+      "loss": 0.021,
+      "step": 1981
+    },
+    {
+      "epoch": 0.39368358327539976,
+      "grad_norm": 0.4596310481079393,
+      "learning_rate": 6.915646393894263e-06,
+      "loss": 0.0126,
+      "step": 1982
+    },
+    {
+      "epoch": 0.3938822127321482,
+      "grad_norm": 0.48189428114732397,
+      "learning_rate": 6.912673985400515e-06,
+      "loss": 0.0119,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3940808421888966,
+      "grad_norm": 0.4738322562132466,
+      "learning_rate": 6.909700784870384e-06,
+      "loss": 0.021,
+      "step": 1984
+    },
+    {
+      "epoch": 0.39427947164564503,
+      "grad_norm": 0.41216128439817645,
+      "learning_rate": 6.906726793535072e-06,
+      "loss": 0.0141,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3944781011023935,
+      "grad_norm": 0.33966794763975494,
+      "learning_rate": 6.903752012626104e-06,
+      "loss": 0.012,
+      "step": 1986
+    },
+    {
+      "epoch": 0.39467673055914193,
+      "grad_norm": 0.2801511238879822,
+      "learning_rate": 6.9007764433753324e-06,
+      "loss": 0.0155,
+      "step": 1987
+    },
+    {
+      "epoch": 0.39487536001589035,
+      "grad_norm": 0.16933671446526258,
+      "learning_rate": 6.897800087014939e-06,
+      "loss": 0.0064,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3950739894726388,
+      "grad_norm": 0.8536198892078616,
+      "learning_rate": 6.894822944777433e-06,
+      "loss": 0.0156,
+      "step": 1989
+    },
+    {
+      "epoch": 0.39527261892938725,
+      "grad_norm": 0.2747101542433025,
+      "learning_rate": 6.891845017895641e-06,
+      "loss": 0.0069,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3954712483861357,
+      "grad_norm": 0.5826546759975959,
+      "learning_rate": 6.888866307602722e-06,
+      "loss": 0.0206,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3956698778428841,
+      "grad_norm": 0.5732473635409778,
+      "learning_rate": 6.885886815132156e-06,
+      "loss": 0.0144,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3958685072996325,
+      "grad_norm": 0.25716654176152753,
+      "learning_rate": 6.882906541717749e-06,
+      "loss": 0.0091,
+      "step": 1993
+    },
+    {
+      "epoch": 0.39606713675638094,
+      "grad_norm": 0.3022114556941632,
+      "learning_rate": 6.879925488593629e-06,
+      "loss": 0.009,
+      "step": 1994
+    },
+    {
+      "epoch": 0.3962657662131294,
+      "grad_norm": 0.4902950717878552,
+      "learning_rate": 6.876943656994246e-06,
+      "loss": 0.0113,
+      "step": 1995
+    },
+    {
+      "epoch": 0.39646439566987784,
+      "grad_norm": 0.46038753699336915,
+      "learning_rate": 6.873961048154374e-06,
+      "loss": 0.0146,
+      "step": 1996
+    },
+    {
+      "epoch": 0.39666302512662627,
+      "grad_norm": 0.6323728442617919,
+      "learning_rate": 6.8709776633091085e-06,
+      "loss": 0.011,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3968616545833747,
+      "grad_norm": 0.4427297026630537,
+      "learning_rate": 6.867993503693868e-06,
+      "loss": 0.0168,
+      "step": 1998
+    },
+    {
+      "epoch": 0.39706028404012317,
+      "grad_norm": 0.4675745693620594,
+      "learning_rate": 6.865008570544387e-06,
+      "loss": 0.0155,
+      "step": 1999
+    },
+    {
+      "epoch": 0.3972589134968716,
+      "grad_norm": 1.0376746039699174,
+      "learning_rate": 6.862022865096724e-06,
+      "loss": 0.0198,
+      "step": 2000
+    },
+    {
+      "epoch": 0.39745754295362,
+      "grad_norm": 0.8009967163692941,
+      "learning_rate": 6.859036388587259e-06,
+      "loss": 0.0097,
+      "step": 2001
+    },
+    {
+      "epoch": 0.39765617241036844,
+      "grad_norm": 0.6394354987730926,
+      "learning_rate": 6.856049142252687e-06,
+      "loss": 0.011,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3978548018671169,
+      "grad_norm": 1.2279342276594911,
+      "learning_rate": 6.853061127330023e-06,
+      "loss": 0.0207,
+      "step": 2003
+    },
+    {
+      "epoch": 0.39805343132386534,
+      "grad_norm": 0.5253473078917413,
+      "learning_rate": 6.850072345056602e-06,
+      "loss": 0.0119,
+      "step": 2004
+    },
+    {
+      "epoch": 0.39825206078061376,
+      "grad_norm": 0.4416957674480141,
+      "learning_rate": 6.847082796670075e-06,
+      "loss": 0.0116,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3984506902373622,
+      "grad_norm": 0.36052345349554943,
+      "learning_rate": 6.844092483408411e-06,
+      "loss": 0.0119,
+      "step": 2006
+    },
+    {
+      "epoch": 0.39864931969411066,
+      "grad_norm": 0.5852938989301314,
+      "learning_rate": 6.8411014065098976e-06,
+      "loss": 0.0155,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3988479491508591,
+      "grad_norm": 0.7611535304666339,
+      "learning_rate": 6.838109567213136e-06,
+      "loss": 0.011,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3990465786076075,
+      "grad_norm": 0.6832215844317394,
+      "learning_rate": 6.835116966757042e-06,
+      "loss": 0.0128,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3992452080643559,
+      "grad_norm": 0.35682225109643834,
+      "learning_rate": 6.832123606380852e-06,
+      "loss": 0.0058,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3994438375211044,
+      "grad_norm": 0.4423053994022838,
+      "learning_rate": 6.82912948732411e-06,
+      "loss": 0.0129,
+      "step": 2011
+    },
+    {
+      "epoch": 0.39964246697785283,
+      "grad_norm": 0.8397841328233963,
+      "learning_rate": 6.826134610826678e-06,
+      "loss": 0.0142,
+      "step": 2012
+    },
+    {
+      "epoch": 0.39984109643460125,
+      "grad_norm": 0.7620187002515791,
+      "learning_rate": 6.8231389781287334e-06,
+      "loss": 0.0191,
+      "step": 2013
+    },
+    {
+      "epoch": 0.4000397258913497,
+      "grad_norm": 0.41386810405085306,
+      "learning_rate": 6.820142590470764e-06,
+      "loss": 0.0129,
+      "step": 2014
+    },
+    {
+      "epoch": 0.4002383553480981,
+      "grad_norm": 0.7141109172857413,
+      "learning_rate": 6.81714544909357e-06,
+      "loss": 0.0183,
+      "step": 2015
+    },
+    {
+      "epoch": 0.4004369848048466,
+      "grad_norm": 0.3552446795117014,
+      "learning_rate": 6.8141475552382665e-06,
+      "loss": 0.0094,
+      "step": 2016
+    },
+    {
+      "epoch": 0.400635614261595,
+      "grad_norm": 0.44775694205320965,
+      "learning_rate": 6.8111489101462755e-06,
+      "loss": 0.0126,
+      "step": 2017
+    },
+    {
+      "epoch": 0.4008342437183434,
+      "grad_norm": 0.5103022028781918,
+      "learning_rate": 6.808149515059336e-06,
+      "loss": 0.015,
+      "step": 2018
+    },
+    {
+      "epoch": 0.40103287317509184,
+      "grad_norm": 1.1270462570715771,
+      "learning_rate": 6.805149371219491e-06,
+      "loss": 0.0214,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4012315026318403,
+      "grad_norm": 0.49338047514807803,
+      "learning_rate": 6.8021484798691005e-06,
+      "loss": 0.0094,
+      "step": 2020
+    },
+    {
+      "epoch": 0.40143013208858874,
+      "grad_norm": 0.4300331782209231,
+      "learning_rate": 6.799146842250829e-06,
+      "loss": 0.0163,
+      "step": 2021
+    },
+    {
+      "epoch": 0.40162876154533717,
+      "grad_norm": 0.33283287964762853,
+      "learning_rate": 6.796144459607652e-06,
+      "loss": 0.0094,
+      "step": 2022
+    },
+    {
+      "epoch": 0.4018273910020856,
+      "grad_norm": 0.3659723339747389,
+      "learning_rate": 6.793141333182853e-06,
+      "loss": 0.0111,
+      "step": 2023
+    },
+    {
+      "epoch": 0.40202602045883407,
+      "grad_norm": 0.49565704935423,
+      "learning_rate": 6.7901374642200235e-06,
+      "loss": 0.0164,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4022246499155825,
+      "grad_norm": 0.6254189475535632,
+      "learning_rate": 6.787132853963063e-06,
+      "loss": 0.0145,
+      "step": 2025
+    },
+    {
+      "epoch": 0.4024232793723309,
+      "grad_norm": 0.39002228377365256,
+      "learning_rate": 6.784127503656177e-06,
+      "loss": 0.0087,
+      "step": 2026
+    },
+    {
+      "epoch": 0.40262190882907933,
+      "grad_norm": 0.2860968118458668,
+      "learning_rate": 6.781121414543878e-06,
+      "loss": 0.0076,
+      "step": 2027
+    },
+    {
+      "epoch": 0.4028205382858278,
+      "grad_norm": 0.37170294728652653,
+      "learning_rate": 6.778114587870985e-06,
+      "loss": 0.0104,
+      "step": 2028
+    },
+    {
+      "epoch": 0.40301916774257623,
+      "grad_norm": 0.4036659979664484,
+      "learning_rate": 6.775107024882623e-06,
+      "loss": 0.007,
+      "step": 2029
+    },
+    {
+      "epoch": 0.40321779719932466,
+      "grad_norm": 0.5549451713316448,
+      "learning_rate": 6.772098726824219e-06,
+      "loss": 0.0119,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4034164266560731,
+      "grad_norm": 0.9605322431392109,
+      "learning_rate": 6.769089694941506e-06,
+      "loss": 0.0165,
+      "step": 2031
+    },
+    {
+      "epoch": 0.40361505611282156,
+      "grad_norm": 0.21818067243379816,
+      "learning_rate": 6.766079930480523e-06,
+      "loss": 0.0055,
+      "step": 2032
+    },
+    {
+      "epoch": 0.40381368556957,
+      "grad_norm": 0.394154840294968,
+      "learning_rate": 6.763069434687608e-06,
+      "loss": 0.0177,
+      "step": 2033
+    },
+    {
+      "epoch": 0.4040123150263184,
+      "grad_norm": 0.4819653179358817,
+      "learning_rate": 6.760058208809407e-06,
+      "loss": 0.0123,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4042109444830668,
+      "grad_norm": 1.280679469593761,
+      "learning_rate": 6.757046254092865e-06,
+      "loss": 0.016,
+      "step": 2035
+    },
+    {
+      "epoch": 0.40440957393981525,
+      "grad_norm": 0.5941803654709766,
+      "learning_rate": 6.754033571785227e-06,
+      "loss": 0.0104,
+      "step": 2036
+    },
+    {
+      "epoch": 0.4046082033965637,
+      "grad_norm": 0.5054580109994721,
+      "learning_rate": 6.7510201631340445e-06,
+      "loss": 0.0165,
+      "step": 2037
+    },
+    {
+      "epoch": 0.40480683285331215,
+      "grad_norm": 0.5427400354352391,
+      "learning_rate": 6.748006029387165e-06,
+      "loss": 0.0102,
+      "step": 2038
+    },
+    {
+      "epoch": 0.40500546231006057,
+      "grad_norm": 0.8181768765515004,
+      "learning_rate": 6.744991171792741e-06,
+      "loss": 0.0104,
+      "step": 2039
+    },
+    {
+      "epoch": 0.405204091766809,
+      "grad_norm": 0.6029402222723133,
+      "learning_rate": 6.741975591599219e-06,
+      "loss": 0.0115,
+      "step": 2040
+    },
+    {
+      "epoch": 0.40540272122355747,
+      "grad_norm": 0.5908122684007071,
+      "learning_rate": 6.738959290055349e-06,
+      "loss": 0.014,
+      "step": 2041
+    },
+    {
+      "epoch": 0.4056013506803059,
+      "grad_norm": 0.2594953161338611,
+      "learning_rate": 6.735942268410178e-06,
+      "loss": 0.009,
+      "step": 2042
+    },
+    {
+      "epoch": 0.4057999801370543,
+      "grad_norm": 0.9374727512612783,
+      "learning_rate": 6.7329245279130525e-06,
+      "loss": 0.0129,
+      "step": 2043
+    },
+    {
+      "epoch": 0.40599860959380274,
+      "grad_norm": 0.978985312338324,
+      "learning_rate": 6.7299060698136146e-06,
+      "loss": 0.0153,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4061972390505512,
+      "grad_norm": 0.4019327020061037,
+      "learning_rate": 6.726886895361807e-06,
+      "loss": 0.014,
+      "step": 2045
+    },
+    {
+      "epoch": 0.40639586850729964,
+      "grad_norm": 0.7831143997306873,
+      "learning_rate": 6.723867005807865e-06,
+      "loss": 0.0185,
+      "step": 2046
+    },
+    {
+      "epoch": 0.40659449796404806,
+      "grad_norm": 0.5998403552907489,
+      "learning_rate": 6.720846402402321e-06,
+      "loss": 0.0133,
+      "step": 2047
+    },
+    {
+      "epoch": 0.4067931274207965,
+      "grad_norm": 0.6325931920573962,
+      "learning_rate": 6.717825086396007e-06,
+      "loss": 0.0091,
+      "step": 2048
+    },
+    {
+      "epoch": 0.40699175687754496,
+      "grad_norm": 1.0061561033266881,
+      "learning_rate": 6.714803059040043e-06,
+      "loss": 0.0142,
+      "step": 2049
+    },
+    {
+      "epoch": 0.4071903863342934,
+      "grad_norm": 0.6985348942673328,
+      "learning_rate": 6.711780321585851e-06,
+      "loss": 0.0118,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4073890157910418,
+      "grad_norm": 1.2385741562172954,
+      "learning_rate": 6.708756875285143e-06,
+      "loss": 0.0112,
+      "step": 2051
+    },
+    {
+      "epoch": 0.40758764524779023,
+      "grad_norm": 0.3580125091627547,
+      "learning_rate": 6.705732721389922e-06,
+      "loss": 0.0089,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4077862747045387,
+      "grad_norm": 0.5168739359216307,
+      "learning_rate": 6.7027078611524895e-06,
+      "loss": 0.012,
+      "step": 2053
+    },
+    {
+      "epoch": 0.40798490416128713,
+      "grad_norm": 1.176391662569905,
+      "learning_rate": 6.699682295825438e-06,
+      "loss": 0.0159,
+      "step": 2054
+    },
+    {
+      "epoch": 0.40818353361803555,
+      "grad_norm": 0.32971423902812974,
+      "learning_rate": 6.69665602666165e-06,
+      "loss": 0.0095,
+      "step": 2055
+    },
+    {
+      "epoch": 0.408382163074784,
+      "grad_norm": 0.7928255269070492,
+      "learning_rate": 6.693629054914298e-06,
+      "loss": 0.0061,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4085807925315324,
+      "grad_norm": 0.35670943004316913,
+      "learning_rate": 6.690601381836852e-06,
+      "loss": 0.0096,
+      "step": 2057
+    },
+    {
+      "epoch": 0.4087794219882809,
+      "grad_norm": 0.7975764075408337,
+      "learning_rate": 6.687573008683067e-06,
+      "loss": 0.015,
+      "step": 2058
+    },
+    {
+      "epoch": 0.4089780514450293,
+      "grad_norm": 0.8319504575889517,
+      "learning_rate": 6.684543936706989e-06,
+      "loss": 0.0094,
+      "step": 2059
+    },
+    {
+      "epoch": 0.4091766809017777,
+      "grad_norm": 0.7205610429728381,
+      "learning_rate": 6.681514167162954e-06,
+      "loss": 0.0217,
+      "step": 2060
+    },
+    {
+      "epoch": 0.40937531035852615,
+      "grad_norm": 0.5035140786062322,
+      "learning_rate": 6.678483701305587e-06,
+      "loss": 0.0143,
+      "step": 2061
+    },
+    {
+      "epoch": 0.4095739398152746,
+      "grad_norm": 0.9123728525009268,
+      "learning_rate": 6.675452540389799e-06,
+      "loss": 0.0146,
+      "step": 2062
+    },
+    {
+      "epoch": 0.40977256927202305,
+      "grad_norm": 1.2095776877970332,
+      "learning_rate": 6.672420685670791e-06,
+      "loss": 0.0232,
+      "step": 2063
+    },
+    {
+      "epoch": 0.40997119872877147,
+      "grad_norm": 0.44951274178922074,
+      "learning_rate": 6.669388138404053e-06,
+      "loss": 0.011,
+      "step": 2064
+    },
+    {
+      "epoch": 0.4101698281855199,
+      "grad_norm": 1.2023015543359805,
+      "learning_rate": 6.666354899845359e-06,
+      "loss": 0.0202,
+      "step": 2065
+    },
+    {
+      "epoch": 0.41036845764226837,
+      "grad_norm": 1.221622773848238,
+      "learning_rate": 6.6633209712507685e-06,
+      "loss": 0.0185,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4105670870990168,
+      "grad_norm": 1.245700627421625,
+      "learning_rate": 6.66028635387663e-06,
+      "loss": 0.0187,
+      "step": 2067
+    },
+    {
+      "epoch": 0.4107657165557652,
+      "grad_norm": 1.3483216195815648,
+      "learning_rate": 6.657251048979576e-06,
+      "loss": 0.0159,
+      "step": 2068
+    },
+    {
+      "epoch": 0.41096434601251364,
+      "grad_norm": 0.782416570047913,
+      "learning_rate": 6.654215057816521e-06,
+      "loss": 0.0172,
+      "step": 2069
+    },
+    {
+      "epoch": 0.4111629754692621,
+      "grad_norm": 0.3826135913453981,
+      "learning_rate": 6.651178381644668e-06,
+      "loss": 0.0147,
+      "step": 2070
+    },
+    {
+      "epoch": 0.41136160492601054,
+      "grad_norm": 0.4239048220757515,
+      "learning_rate": 6.648141021721499e-06,
+      "loss": 0.0131,
+      "step": 2071
+    },
+    {
+      "epoch": 0.41156023438275896,
+      "grad_norm": 0.20186798617854174,
+      "learning_rate": 6.645102979304785e-06,
+      "loss": 0.0067,
+      "step": 2072
+    },
+    {
+      "epoch": 0.4117588638395074,
+      "grad_norm": 0.44996535792808406,
+      "learning_rate": 6.642064255652576e-06,
+      "loss": 0.0134,
+      "step": 2073
+    },
+    {
+      "epoch": 0.41195749329625586,
+      "grad_norm": 0.7266208044448531,
+      "learning_rate": 6.6390248520232005e-06,
+      "loss": 0.0104,
+      "step": 2074
+    },
+    {
+      "epoch": 0.4121561227530043,
+      "grad_norm": 0.8966072482079172,
+      "learning_rate": 6.6359847696752765e-06,
+      "loss": 0.019,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4123547522097527,
+      "grad_norm": 0.39163729067805625,
+      "learning_rate": 6.6329440098676975e-06,
+      "loss": 0.0125,
+      "step": 2076
+    },
+    {
+      "epoch": 0.41255338166650113,
+      "grad_norm": 0.5443082189044185,
+      "learning_rate": 6.629902573859639e-06,
+      "loss": 0.0127,
+      "step": 2077
+    },
+    {
+      "epoch": 0.41275201112324955,
+      "grad_norm": 0.8156259321153082,
+      "learning_rate": 6.626860462910557e-06,
+      "loss": 0.0175,
+      "step": 2078
+    },
+    {
+      "epoch": 0.41295064057999803,
+      "grad_norm": 0.7619649973471931,
+      "learning_rate": 6.623817678280187e-06,
+      "loss": 0.0136,
+      "step": 2079
+    },
+    {
+      "epoch": 0.41314927003674645,
+      "grad_norm": 0.779374490528315,
+      "learning_rate": 6.620774221228542e-06,
+      "loss": 0.0155,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4133478994934949,
+      "grad_norm": 0.3046402817702944,
+      "learning_rate": 6.617730093015915e-06,
+      "loss": 0.0108,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4135465289502433,
+      "grad_norm": 0.442045865809681,
+      "learning_rate": 6.614685294902876e-06,
+      "loss": 0.0121,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4137451584069918,
+      "grad_norm": 0.5778707285503455,
+      "learning_rate": 6.611639828150273e-06,
+      "loss": 0.0133,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4139437878637402,
+      "grad_norm": 0.8172818474796,
+      "learning_rate": 6.608593694019233e-06,
+      "loss": 0.0118,
+      "step": 2084
+    },
+    {
+      "epoch": 0.4141424173204886,
+      "grad_norm": 1.0008201942397041,
+      "learning_rate": 6.605546893771155e-06,
+      "loss": 0.0099,
+      "step": 2085
+    },
+    {
+      "epoch": 0.41434104677723704,
+      "grad_norm": 0.32404676420516487,
+      "learning_rate": 6.602499428667717e-06,
+      "loss": 0.008,
+      "step": 2086
+    },
+    {
+      "epoch": 0.4145396762339855,
+      "grad_norm": 0.9189946664738823,
+      "learning_rate": 6.599451299970867e-06,
+      "loss": 0.0102,
+      "step": 2087
+    },
+    {
+      "epoch": 0.41473830569073394,
+      "grad_norm": 0.5740526296204994,
+      "learning_rate": 6.596402508942838e-06,
+      "loss": 0.0186,
+      "step": 2088
+    },
+    {
+      "epoch": 0.41493693514748237,
+      "grad_norm": 0.4517267054867236,
+      "learning_rate": 6.593353056846128e-06,
+      "loss": 0.017,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4151355646042308,
+      "grad_norm": 0.6912662883562258,
+      "learning_rate": 6.590302944943513e-06,
+      "loss": 0.0149,
+      "step": 2090
+    },
+    {
+      "epoch": 0.41533419406097927,
+      "grad_norm": 0.7887240811333215,
+      "learning_rate": 6.58725217449804e-06,
+      "loss": 0.0118,
+      "step": 2091
+    },
+    {
+      "epoch": 0.4155328235177277,
+      "grad_norm": 1.0223004745638395,
+      "learning_rate": 6.584200746773033e-06,
+      "loss": 0.0222,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4157314529744761,
+      "grad_norm": 0.4963546873514082,
+      "learning_rate": 6.581148663032082e-06,
+      "loss": 0.0154,
+      "step": 2093
+    },
+    {
+      "epoch": 0.41593008243122453,
+      "grad_norm": 0.9897376857036329,
+      "learning_rate": 6.5780959245390516e-06,
+      "loss": 0.0132,
+      "step": 2094
+    },
+    {
+      "epoch": 0.416128711887973,
+      "grad_norm": 0.5755955971080043,
+      "learning_rate": 6.5750425325580815e-06,
+      "loss": 0.0164,
+      "step": 2095
+    },
+    {
+      "epoch": 0.41632734134472144,
+      "grad_norm": 0.4895034942702204,
+      "learning_rate": 6.571988488353574e-06,
+      "loss": 0.0096,
+      "step": 2096
+    },
+    {
+      "epoch": 0.41652597080146986,
+      "grad_norm": 0.8734463704411604,
+      "learning_rate": 6.5689337931902106e-06,
+      "loss": 0.0139,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4167246002582183,
+      "grad_norm": 0.3329657332391327,
+      "learning_rate": 6.565878448332933e-06,
+      "loss": 0.0093,
+      "step": 2098
+    },
+    {
+      "epoch": 0.4169232297149667,
+      "grad_norm": 0.40090511983879706,
+      "learning_rate": 6.562822455046958e-06,
+      "loss": 0.0076,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4171218591717152,
+      "grad_norm": 1.05951571631075,
+      "learning_rate": 6.559765814597768e-06,
+      "loss": 0.0173,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4173204886284636,
+      "grad_norm": 0.7267645054655884,
+      "learning_rate": 6.556708528251119e-06,
+      "loss": 0.0166,
+      "step": 2101
+    },
+    {
+      "epoch": 0.417519118085212,
+      "grad_norm": 0.4476817973691744,
+      "learning_rate": 6.553650597273025e-06,
+      "loss": 0.0098,
+      "step": 2102
+    },
+    {
+      "epoch": 0.41771774754196045,
+      "grad_norm": 0.8482543101543035,
+      "learning_rate": 6.550592022929776e-06,
+      "loss": 0.0206,
+      "step": 2103
+    },
+    {
+      "epoch": 0.4179163769987089,
+      "grad_norm": 0.3610267762233538,
+      "learning_rate": 6.5475328064879225e-06,
+      "loss": 0.0103,
+      "step": 2104
+    },
+    {
+      "epoch": 0.41811500645545735,
+      "grad_norm": 0.8566164925435632,
+      "learning_rate": 6.544472949214283e-06,
+      "loss": 0.0127,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4183136359122058,
+      "grad_norm": 0.33243771966184194,
+      "learning_rate": 6.5414124523759425e-06,
+      "loss": 0.0098,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4185122653689542,
+      "grad_norm": 0.380868671779114,
+      "learning_rate": 6.538351317240249e-06,
+      "loss": 0.0131,
+      "step": 2107
+    },
+    {
+      "epoch": 0.4187108948257027,
+      "grad_norm": 0.5135335899779155,
+      "learning_rate": 6.5352895450748155e-06,
+      "loss": 0.0067,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4189095242824511,
+      "grad_norm": 0.38253084605058807,
+      "learning_rate": 6.53222713714752e-06,
+      "loss": 0.0096,
+      "step": 2109
+    },
+    {
+      "epoch": 0.4191081537391995,
+      "grad_norm": 0.6100179434270661,
+      "learning_rate": 6.5291640947265015e-06,
+      "loss": 0.0116,
+      "step": 2110
+    },
+    {
+      "epoch": 0.41930678319594794,
+      "grad_norm": 1.412550135096153,
+      "learning_rate": 6.5261004190801615e-06,
+      "loss": 0.0201,
+      "step": 2111
+    },
+    {
+      "epoch": 0.4195054126526964,
+      "grad_norm": 0.26253036733279284,
+      "learning_rate": 6.523036111477165e-06,
+      "loss": 0.0057,
+      "step": 2112
+    },
+    {
+      "epoch": 0.41970404210944484,
+      "grad_norm": 0.7640094825754855,
+      "learning_rate": 6.519971173186441e-06,
+      "loss": 0.0129,
+      "step": 2113
+    },
+    {
+      "epoch": 0.41990267156619326,
+      "grad_norm": 0.41106174134380463,
+      "learning_rate": 6.516905605477177e-06,
+      "loss": 0.0081,
+      "step": 2114
+    },
+    {
+      "epoch": 0.4201013010229417,
+      "grad_norm": 0.45933590667057456,
+      "learning_rate": 6.513839409618821e-06,
+      "loss": 0.0078,
+      "step": 2115
+    },
+    {
+      "epoch": 0.42029993047969016,
+      "grad_norm": 0.7965887185750786,
+      "learning_rate": 6.510772586881081e-06,
+      "loss": 0.0179,
+      "step": 2116
+    },
+    {
+      "epoch": 0.4204985599364386,
+      "grad_norm": 0.59035117918782,
+      "learning_rate": 6.5077051385339266e-06,
+      "loss": 0.0096,
+      "step": 2117
+    },
+    {
+      "epoch": 0.420697189393187,
+      "grad_norm": 0.2143236615237929,
+      "learning_rate": 6.504637065847583e-06,
+      "loss": 0.0045,
+      "step": 2118
+    },
+    {
+      "epoch": 0.42089581884993543,
+      "grad_norm": 0.7334121261956231,
+      "learning_rate": 6.501568370092536e-06,
+      "loss": 0.0147,
+      "step": 2119
+    },
+    {
+      "epoch": 0.42109444830668386,
+      "grad_norm": 0.7597269250272318,
+      "learning_rate": 6.498499052539533e-06,
+      "loss": 0.0092,
+      "step": 2120
+    },
+    {
+      "epoch": 0.42129307776343233,
+      "grad_norm": 0.47333533855501414,
+      "learning_rate": 6.495429114459573e-06,
+      "loss": 0.0065,
+      "step": 2121
+    },
+    {
+      "epoch": 0.42149170722018076,
+      "grad_norm": 1.330156731921058,
+      "learning_rate": 6.4923585571239135e-06,
+      "loss": 0.0182,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4216903366769292,
+      "grad_norm": 0.6948668654706006,
+      "learning_rate": 6.48928738180407e-06,
+      "loss": 0.0085,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4218889661336776,
+      "grad_norm": 0.7890280489148934,
+      "learning_rate": 6.486215589771812e-06,
+      "loss": 0.015,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4220875955904261,
+      "grad_norm": 0.452242433827678,
+      "learning_rate": 6.483143182299168e-06,
+      "loss": 0.0051,
+      "step": 2125
+    },
+    {
+      "epoch": 0.4222862250471745,
+      "grad_norm": 0.5748220685978489,
+      "learning_rate": 6.480070160658416e-06,
+      "loss": 0.0185,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4224848545039229,
+      "grad_norm": 1.0878583524025376,
+      "learning_rate": 6.476996526122095e-06,
+      "loss": 0.0176,
+      "step": 2127
+    },
+    {
+      "epoch": 0.42268348396067135,
+      "grad_norm": 1.0214588977314998,
+      "learning_rate": 6.47392227996299e-06,
+      "loss": 0.0229,
+      "step": 2128
+    },
+    {
+      "epoch": 0.4228821134174198,
+      "grad_norm": 0.8368364297826585,
+      "learning_rate": 6.470847423454145e-06,
+      "loss": 0.0119,
+      "step": 2129
+    },
+    {
+      "epoch": 0.42308074287416825,
+      "grad_norm": 1.5630364581414389,
+      "learning_rate": 6.467771957868856e-06,
+      "loss": 0.0279,
+      "step": 2130
+    },
+    {
+      "epoch": 0.42327937233091667,
+      "grad_norm": 0.5265919309050093,
+      "learning_rate": 6.464695884480666e-06,
+      "loss": 0.0066,
+      "step": 2131
+    },
+    {
+      "epoch": 0.4234780017876651,
+      "grad_norm": 0.5458391814360123,
+      "learning_rate": 6.461619204563379e-06,
+      "loss": 0.0078,
+      "step": 2132
+    },
+    {
+      "epoch": 0.42367663124441357,
+      "grad_norm": 0.5629575812962587,
+      "learning_rate": 6.458541919391046e-06,
+      "loss": 0.0164,
+      "step": 2133
+    },
+    {
+      "epoch": 0.423875260701162,
+      "grad_norm": 1.0217593729153478,
+      "learning_rate": 6.455464030237961e-06,
+      "loss": 0.0132,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4240738901579104,
+      "grad_norm": 0.3745957899282288,
+      "learning_rate": 6.452385538378681e-06,
+      "loss": 0.0078,
+      "step": 2135
+    },
+    {
+      "epoch": 0.42427251961465884,
+      "grad_norm": 0.4083516790878554,
+      "learning_rate": 6.4493064450880014e-06,
+      "loss": 0.0102,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4244711490714073,
+      "grad_norm": 1.0042440949633358,
+      "learning_rate": 6.446226751640976e-06,
+      "loss": 0.0146,
+      "step": 2137
+    },
+    {
+      "epoch": 0.42466977852815574,
+      "grad_norm": 0.6514777945792959,
+      "learning_rate": 6.4431464593129e-06,
+      "loss": 0.0108,
+      "step": 2138
+    },
+    {
+      "epoch": 0.42486840798490416,
+      "grad_norm": 0.8367257022237136,
+      "learning_rate": 6.440065569379321e-06,
+      "loss": 0.0221,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4250670374416526,
+      "grad_norm": 1.226693427969879,
+      "learning_rate": 6.436984083116032e-06,
+      "loss": 0.0229,
+      "step": 2140
+    },
+    {
+      "epoch": 0.425265666898401,
+      "grad_norm": 0.5627979518455061,
+      "learning_rate": 6.433902001799074e-06,
+      "loss": 0.012,
+      "step": 2141
+    },
+    {
+      "epoch": 0.4254642963551495,
+      "grad_norm": 0.7446434572287394,
+      "learning_rate": 6.430819326704732e-06,
+      "loss": 0.0133,
+      "step": 2142
+    },
+    {
+      "epoch": 0.4256629258118979,
+      "grad_norm": 0.5119208241159949,
+      "learning_rate": 6.427736059109539e-06,
+      "loss": 0.0134,
+      "step": 2143
+    },
+    {
+      "epoch": 0.42586155526864633,
+      "grad_norm": 0.43798331629976245,
+      "learning_rate": 6.424652200290275e-06,
+      "loss": 0.0116,
+      "step": 2144
+    },
+    {
+      "epoch": 0.42606018472539475,
+      "grad_norm": 0.8345104818237911,
+      "learning_rate": 6.421567751523962e-06,
+      "loss": 0.0146,
+      "step": 2145
+    },
+    {
+      "epoch": 0.42625881418214323,
+      "grad_norm": 0.5168528918687458,
+      "learning_rate": 6.4184827140878655e-06,
+      "loss": 0.0164,
+      "step": 2146
+    },
+    {
+      "epoch": 0.42645744363889165,
+      "grad_norm": 0.769263430644991,
+      "learning_rate": 6.415397089259497e-06,
+      "loss": 0.011,
+      "step": 2147
+    },
+    {
+      "epoch": 0.4266560730956401,
+      "grad_norm": 1.2006056414732815,
+      "learning_rate": 6.412310878316612e-06,
+      "loss": 0.0199,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4268547025523885,
+      "grad_norm": 0.36140757435107507,
+      "learning_rate": 6.409224082537206e-06,
+      "loss": 0.0095,
+      "step": 2149
+    },
+    {
+      "epoch": 0.427053332009137,
+      "grad_norm": 0.4887806589406039,
+      "learning_rate": 6.4061367031995205e-06,
+      "loss": 0.0137,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4272519614658854,
+      "grad_norm": 0.4372668846370759,
+      "learning_rate": 6.403048741582033e-06,
+      "loss": 0.0143,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4274505909226338,
+      "grad_norm": 0.4264517517122641,
+      "learning_rate": 6.3999601989634665e-06,
+      "loss": 0.011,
+      "step": 2152
+    },
+    {
+      "epoch": 0.42764922037938224,
+      "grad_norm": 0.5939697458462438,
+      "learning_rate": 6.396871076622784e-06,
+      "loss": 0.0153,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4278478498361307,
+      "grad_norm": 0.8720261216914841,
+      "learning_rate": 6.393781375839187e-06,
+      "loss": 0.0133,
+      "step": 2154
+    },
+    {
+      "epoch": 0.42804647929287915,
+      "grad_norm": 0.39205208543034786,
+      "learning_rate": 6.390691097892119e-06,
+      "loss": 0.0122,
+      "step": 2155
+    },
+    {
+      "epoch": 0.42824510874962757,
+      "grad_norm": 0.3140641676904937,
+      "learning_rate": 6.387600244061258e-06,
+      "loss": 0.0074,
+      "step": 2156
+    },
+    {
+      "epoch": 0.428443738206376,
+      "grad_norm": 0.5950724392389523,
+      "learning_rate": 6.384508815626529e-06,
+      "loss": 0.0119,
+      "step": 2157
+    },
+    {
+      "epoch": 0.42864236766312447,
+      "grad_norm": 1.1164418050217393,
+      "learning_rate": 6.381416813868085e-06,
+      "loss": 0.0168,
+      "step": 2158
+    },
+    {
+      "epoch": 0.4288409971198729,
+      "grad_norm": 1.0956077716877128,
+      "learning_rate": 6.37832424006632e-06,
+      "loss": 0.0246,
+      "step": 2159
+    },
+    {
+      "epoch": 0.4290396265766213,
+      "grad_norm": 0.5788002270708077,
+      "learning_rate": 6.375231095501868e-06,
+      "loss": 0.016,
+      "step": 2160
+    },
+    {
+      "epoch": 0.42923825603336974,
+      "grad_norm": 0.634313578842633,
+      "learning_rate": 6.3721373814555965e-06,
+      "loss": 0.0083,
+      "step": 2161
+    },
+    {
+      "epoch": 0.42943688549011816,
+      "grad_norm": 0.3256946524279106,
+      "learning_rate": 6.369043099208609e-06,
+      "loss": 0.0151,
+      "step": 2162
+    },
+    {
+      "epoch": 0.42963551494686664,
+      "grad_norm": 1.4280037849128235,
+      "learning_rate": 6.365948250042246e-06,
+      "loss": 0.0127,
+      "step": 2163
+    },
+    {
+      "epoch": 0.42983414440361506,
+      "grad_norm": 0.34022719740651086,
+      "learning_rate": 6.3628528352380804e-06,
+      "loss": 0.0073,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4300327738603635,
+      "grad_norm": 0.5438789476793532,
+      "learning_rate": 6.359756856077921e-06,
+      "loss": 0.0109,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4302314033171119,
+      "grad_norm": 0.9150416889469685,
+      "learning_rate": 6.3566603138438074e-06,
+      "loss": 0.0176,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4304300327738604,
+      "grad_norm": 0.37232770382572944,
+      "learning_rate": 6.353563209818019e-06,
+      "loss": 0.0108,
+      "step": 2167
+    },
+    {
+      "epoch": 0.4306286622306088,
+      "grad_norm": 0.5476731401751236,
+      "learning_rate": 6.35046554528306e-06,
+      "loss": 0.0097,
+      "step": 2168
+    },
+    {
+      "epoch": 0.43082729168735723,
+      "grad_norm": 0.996056551510471,
+      "learning_rate": 6.347367321521671e-06,
+      "loss": 0.0164,
+      "step": 2169
+    },
+    {
+      "epoch": 0.43102592114410565,
+      "grad_norm": 0.7735149654496623,
+      "learning_rate": 6.3442685398168226e-06,
+      "loss": 0.0174,
+      "step": 2170
+    },
+    {
+      "epoch": 0.43122455060085413,
+      "grad_norm": 0.5452867098400445,
+      "learning_rate": 6.3411692014517175e-06,
+      "loss": 0.0129,
+      "step": 2171
+    },
+    {
+      "epoch": 0.43142318005760255,
+      "grad_norm": 0.4948507173958259,
+      "learning_rate": 6.338069307709791e-06,
+      "loss": 0.0154,
+      "step": 2172
+    },
+    {
+      "epoch": 0.431621809514351,
+      "grad_norm": 0.612880978569296,
+      "learning_rate": 6.334968859874704e-06,
+      "loss": 0.0177,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4318204389710994,
+      "grad_norm": 0.6349394426674493,
+      "learning_rate": 6.331867859230347e-06,
+      "loss": 0.0123,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4320190684278479,
+      "grad_norm": 0.44612232259436474,
+      "learning_rate": 6.328766307060845e-06,
+      "loss": 0.015,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4322176978845963,
+      "grad_norm": 0.8671464943443922,
+      "learning_rate": 6.3256642046505455e-06,
+      "loss": 0.0155,
+      "step": 2176
+    },
+    {
+      "epoch": 0.4324163273413447,
+      "grad_norm": 0.35014656407021144,
+      "learning_rate": 6.322561553284027e-06,
+      "loss": 0.0049,
+      "step": 2177
+    },
+    {
+      "epoch": 0.43261495679809314,
+      "grad_norm": 1.0746580549448301,
+      "learning_rate": 6.319458354246093e-06,
+      "loss": 0.0128,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4328135862548416,
+      "grad_norm": 0.7495218834739403,
+      "learning_rate": 6.316354608821775e-06,
+      "loss": 0.0167,
+      "step": 2179
+    },
+    {
+      "epoch": 0.43301221571159004,
+      "grad_norm": 0.6790113545596657,
+      "learning_rate": 6.313250318296333e-06,
+      "loss": 0.0071,
+      "step": 2180
+    },
+    {
+      "epoch": 0.43321084516833847,
+      "grad_norm": 0.5279299100486009,
+      "learning_rate": 6.31014548395525e-06,
+      "loss": 0.0125,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4334094746250869,
+      "grad_norm": 0.3722980884712896,
+      "learning_rate": 6.307040107084232e-06,
+      "loss": 0.0067,
+      "step": 2182
+    },
+    {
+      "epoch": 0.4336081040818353,
+      "grad_norm": 0.5681920025290187,
+      "learning_rate": 6.3039341889692164e-06,
+      "loss": 0.0168,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4338067335385838,
+      "grad_norm": 0.7370282183909415,
+      "learning_rate": 6.300827730896359e-06,
+      "loss": 0.013,
+      "step": 2184
+    },
+    {
+      "epoch": 0.4340053629953322,
+      "grad_norm": 1.0952586557587631,
+      "learning_rate": 6.29772073415204e-06,
+      "loss": 0.0246,
+      "step": 2185
+    },
+    {
+      "epoch": 0.43420399245208063,
+      "grad_norm": 0.36852862876653647,
+      "learning_rate": 6.294613200022865e-06,
+      "loss": 0.0073,
+      "step": 2186
+    },
+    {
+      "epoch": 0.43440262190882906,
+      "grad_norm": 0.5506267036624556,
+      "learning_rate": 6.291505129795659e-06,
+      "loss": 0.0078,
+      "step": 2187
+    },
+    {
+      "epoch": 0.43460125136557753,
+      "grad_norm": 0.642929767010988,
+      "learning_rate": 6.288396524757473e-06,
+      "loss": 0.02,
+      "step": 2188
+    },
+    {
+      "epoch": 0.43479988082232596,
+      "grad_norm": 0.4663773137832936,
+      "learning_rate": 6.285287386195577e-06,
+      "loss": 0.0096,
+      "step": 2189
+    },
+    {
+      "epoch": 0.4349985102790744,
+      "grad_norm": 0.899128172317911,
+      "learning_rate": 6.28217771539746e-06,
+      "loss": 0.0165,
+      "step": 2190
+    },
+    {
+      "epoch": 0.4351971397358228,
+      "grad_norm": 0.5828032708204945,
+      "learning_rate": 6.279067513650837e-06,
+      "loss": 0.0138,
+      "step": 2191
+    },
+    {
+      "epoch": 0.4353957691925713,
+      "grad_norm": 0.5191364681793799,
+      "learning_rate": 6.275956782243637e-06,
+      "loss": 0.0117,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4355943986493197,
+      "grad_norm": 0.5129148486056785,
+      "learning_rate": 6.272845522464011e-06,
+      "loss": 0.0125,
+      "step": 2193
+    },
+    {
+      "epoch": 0.4357930281060681,
+      "grad_norm": 0.5310527010612295,
+      "learning_rate": 6.269733735600328e-06,
+      "loss": 0.0106,
+      "step": 2194
+    },
+    {
+      "epoch": 0.43599165756281655,
+      "grad_norm": 0.4310848203884204,
+      "learning_rate": 6.266621422941177e-06,
+      "loss": 0.0124,
+      "step": 2195
+    },
+    {
+      "epoch": 0.436190287019565,
+      "grad_norm": 0.9514944377133658,
+      "learning_rate": 6.263508585775364e-06,
+      "loss": 0.0127,
+      "step": 2196
+    },
+    {
+      "epoch": 0.43638891647631345,
+      "grad_norm": 0.7287942888398268,
+      "learning_rate": 6.260395225391911e-06,
+      "loss": 0.0107,
+      "step": 2197
+    },
+    {
+      "epoch": 0.43658754593306187,
+      "grad_norm": 0.3824710094767856,
+      "learning_rate": 6.2572813430800565e-06,
+      "loss": 0.0131,
+      "step": 2198
+    },
+    {
+      "epoch": 0.4367861753898103,
+      "grad_norm": 0.4588648637376047,
+      "learning_rate": 6.254166940129256e-06,
+      "loss": 0.0152,
+      "step": 2199
+    },
+    {
+      "epoch": 0.43698480484655877,
+      "grad_norm": 0.7375069280141057,
+      "learning_rate": 6.251052017829183e-06,
+      "loss": 0.0132,
+      "step": 2200
+    },
+    {
+      "epoch": 0.4371834343033072,
+      "grad_norm": 0.6624079769881998,
+      "learning_rate": 6.247936577469722e-06,
+      "loss": 0.0131,
+      "step": 2201
+    },
+    {
+      "epoch": 0.4373820637600556,
+      "grad_norm": 0.2923891307524442,
+      "learning_rate": 6.244820620340974e-06,
+      "loss": 0.0098,
+      "step": 2202
+    },
+    {
+      "epoch": 0.43758069321680404,
+      "grad_norm": 1.2709855509660637,
+      "learning_rate": 6.241704147733254e-06,
+      "loss": 0.0213,
+      "step": 2203
+    },
+    {
+      "epoch": 0.43777932267355246,
+      "grad_norm": 0.593801147450338,
+      "learning_rate": 6.2385871609370885e-06,
+      "loss": 0.0172,
+      "step": 2204
+    },
+    {
+      "epoch": 0.43797795213030094,
+      "grad_norm": 0.9452658486066292,
+      "learning_rate": 6.235469661243222e-06,
+      "loss": 0.0191,
+      "step": 2205
+    },
+    {
+      "epoch": 0.43817658158704936,
+      "grad_norm": 0.46873430033477437,
+      "learning_rate": 6.2323516499426055e-06,
+      "loss": 0.017,
+      "step": 2206
+    },
+    {
+      "epoch": 0.4383752110437978,
+      "grad_norm": 0.934149350494909,
+      "learning_rate": 6.229233128326404e-06,
+      "loss": 0.0168,
+      "step": 2207
+    },
+    {
+      "epoch": 0.4385738405005462,
+      "grad_norm": 0.7694849628244852,
+      "learning_rate": 6.226114097685996e-06,
+      "loss": 0.0106,
+      "step": 2208
+    },
+    {
+      "epoch": 0.4387724699572947,
+      "grad_norm": 0.36801759673916484,
+      "learning_rate": 6.222994559312967e-06,
+      "loss": 0.0058,
+      "step": 2209
+    },
+    {
+      "epoch": 0.4389710994140431,
+      "grad_norm": 0.6695445746957661,
+      "learning_rate": 6.219874514499116e-06,
+      "loss": 0.0267,
+      "step": 2210
+    },
+    {
+      "epoch": 0.43916972887079153,
+      "grad_norm": 0.38996735486229944,
+      "learning_rate": 6.21675396453645e-06,
+      "loss": 0.009,
+      "step": 2211
+    },
+    {
+      "epoch": 0.43936835832753995,
+      "grad_norm": 0.25082029907152686,
+      "learning_rate": 6.213632910717186e-06,
+      "loss": 0.0089,
+      "step": 2212
+    },
+    {
+      "epoch": 0.43956698778428843,
+      "grad_norm": 0.30323741324880354,
+      "learning_rate": 6.210511354333749e-06,
+      "loss": 0.0114,
+      "step": 2213
+    },
+    {
+      "epoch": 0.43976561724103685,
+      "grad_norm": 0.611598681425317,
+      "learning_rate": 6.207389296678773e-06,
+      "loss": 0.0237,
+      "step": 2214
+    },
+    {
+      "epoch": 0.4399642466977853,
+      "grad_norm": 0.5733866943683243,
+      "learning_rate": 6.2042667390450975e-06,
+      "loss": 0.0097,
+      "step": 2215
+    },
+    {
+      "epoch": 0.4401628761545337,
+      "grad_norm": 0.79670133428446,
+      "learning_rate": 6.201143682725772e-06,
+      "loss": 0.0176,
+      "step": 2216
+    },
+    {
+      "epoch": 0.4403615056112822,
+      "grad_norm": 0.9551481409610042,
+      "learning_rate": 6.19802012901405e-06,
+      "loss": 0.0143,
+      "step": 2217
+    },
+    {
+      "epoch": 0.4405601350680306,
+      "grad_norm": 0.5431923190522518,
+      "learning_rate": 6.1948960792033926e-06,
+      "loss": 0.0231,
+      "step": 2218
+    },
+    {
+      "epoch": 0.440758764524779,
+      "grad_norm": 1.0130427924002454,
+      "learning_rate": 6.191771534587465e-06,
+      "loss": 0.0125,
+      "step": 2219
+    },
+    {
+      "epoch": 0.44095739398152745,
+      "grad_norm": 0.5857570298238707,
+      "learning_rate": 6.188646496460139e-06,
+      "loss": 0.0135,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4411560234382759,
+      "grad_norm": 0.9795520817741151,
+      "learning_rate": 6.185520966115489e-06,
+      "loss": 0.0154,
+      "step": 2221
+    },
+    {
+      "epoch": 0.44135465289502435,
+      "grad_norm": 0.7885367572135659,
+      "learning_rate": 6.182394944847794e-06,
+      "loss": 0.0173,
+      "step": 2222
+    },
+    {
+      "epoch": 0.44155328235177277,
+      "grad_norm": 0.21893643286317785,
+      "learning_rate": 6.179268433951536e-06,
+      "loss": 0.0058,
+      "step": 2223
+    },
+    {
+      "epoch": 0.4417519118085212,
+      "grad_norm": 0.7459506265711627,
+      "learning_rate": 6.1761414347213995e-06,
+      "loss": 0.0089,
+      "step": 2224
+    },
+    {
+      "epoch": 0.4419505412652696,
+      "grad_norm": 0.27792088533372517,
+      "learning_rate": 6.173013948452274e-06,
+      "loss": 0.0098,
+      "step": 2225
+    },
+    {
+      "epoch": 0.4421491707220181,
+      "grad_norm": 1.2345740081065522,
+      "learning_rate": 6.1698859764392475e-06,
+      "loss": 0.0143,
+      "step": 2226
+    },
+    {
+      "epoch": 0.4423478001787665,
+      "grad_norm": 0.3732015361677137,
+      "learning_rate": 6.1667575199776096e-06,
+      "loss": 0.0131,
+      "step": 2227
+    },
+    {
+      "epoch": 0.44254642963551494,
+      "grad_norm": 0.5343008204160175,
+      "learning_rate": 6.163628580362851e-06,
+      "loss": 0.0089,
+      "step": 2228
+    },
+    {
+      "epoch": 0.44274505909226336,
+      "grad_norm": 0.5522418996318227,
+      "learning_rate": 6.160499158890664e-06,
+      "loss": 0.0106,
+      "step": 2229
+    },
+    {
+      "epoch": 0.44294368854901184,
+      "grad_norm": 0.6254354935039006,
+      "learning_rate": 6.157369256856934e-06,
+      "loss": 0.0167,
+      "step": 2230
+    },
+    {
+      "epoch": 0.44314231800576026,
+      "grad_norm": 0.42379511270760073,
+      "learning_rate": 6.154238875557755e-06,
+      "loss": 0.0131,
+      "step": 2231
+    },
+    {
+      "epoch": 0.4433409474625087,
+      "grad_norm": 0.6058120408697395,
+      "learning_rate": 6.151108016289416e-06,
+      "loss": 0.0096,
+      "step": 2232
+    },
+    {
+      "epoch": 0.4435395769192571,
+      "grad_norm": 0.25542634551541554,
+      "learning_rate": 6.147976680348398e-06,
+      "loss": 0.0072,
+      "step": 2233
+    },
+    {
+      "epoch": 0.4437382063760056,
+      "grad_norm": 0.2899452724247626,
+      "learning_rate": 6.144844869031385e-06,
+      "loss": 0.006,
+      "step": 2234
+    },
+    {
+      "epoch": 0.443936835832754,
+      "grad_norm": 0.7491935474647948,
+      "learning_rate": 6.141712583635261e-06,
+      "loss": 0.0145,
+      "step": 2235
+    },
+    {
+      "epoch": 0.44413546528950243,
+      "grad_norm": 0.4422482327524406,
+      "learning_rate": 6.138579825457098e-06,
+      "loss": 0.0092,
+      "step": 2236
+    },
+    {
+      "epoch": 0.44433409474625085,
+      "grad_norm": 0.21771092100081543,
+      "learning_rate": 6.135446595794169e-06,
+      "loss": 0.0085,
+      "step": 2237
+    },
+    {
+      "epoch": 0.44453272420299933,
+      "grad_norm": 1.4002752559333222,
+      "learning_rate": 6.132312895943942e-06,
+      "loss": 0.0207,
+      "step": 2238
+    },
+    {
+      "epoch": 0.44473135365974775,
+      "grad_norm": 0.7190190028304118,
+      "learning_rate": 6.129178727204079e-06,
+      "loss": 0.0161,
+      "step": 2239
+    },
+    {
+      "epoch": 0.4449299831164962,
+      "grad_norm": 0.42822061766771496,
+      "learning_rate": 6.126044090872437e-06,
+      "loss": 0.0124,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4451286125732446,
+      "grad_norm": 0.677711220443176,
+      "learning_rate": 6.1229089882470645e-06,
+      "loss": 0.0101,
+      "step": 2241
+    },
+    {
+      "epoch": 0.4453272420299931,
+      "grad_norm": 1.059203479885282,
+      "learning_rate": 6.119773420626203e-06,
+      "loss": 0.0325,
+      "step": 2242
+    },
+    {
+      "epoch": 0.4455258714867415,
+      "grad_norm": 0.5706231879648851,
+      "learning_rate": 6.1166373893082895e-06,
+      "loss": 0.0106,
+      "step": 2243
+    },
+    {
+      "epoch": 0.4457245009434899,
+      "grad_norm": 0.995491363379659,
+      "learning_rate": 6.113500895591953e-06,
+      "loss": 0.0167,
+      "step": 2244
+    },
+    {
+      "epoch": 0.44592313040023834,
+      "grad_norm": 0.29779451432903564,
+      "learning_rate": 6.110363940776008e-06,
+      "loss": 0.0092,
+      "step": 2245
+    },
+    {
+      "epoch": 0.44612175985698677,
+      "grad_norm": 0.5762002893389644,
+      "learning_rate": 6.10722652615947e-06,
+      "loss": 0.0143,
+      "step": 2246
+    },
+    {
+      "epoch": 0.44632038931373524,
+      "grad_norm": 0.7215045066061389,
+      "learning_rate": 6.104088653041534e-06,
+      "loss": 0.0174,
+      "step": 2247
+    },
+    {
+      "epoch": 0.44651901877048367,
+      "grad_norm": 0.32513640970058094,
+      "learning_rate": 6.100950322721594e-06,
+      "loss": 0.0041,
+      "step": 2248
+    },
+    {
+      "epoch": 0.4467176482272321,
+      "grad_norm": 0.5763190265316815,
+      "learning_rate": 6.097811536499227e-06,
+      "loss": 0.018,
+      "step": 2249
+    },
+    {
+      "epoch": 0.4469162776839805,
+      "grad_norm": 0.8055544487235922,
+      "learning_rate": 6.094672295674202e-06,
+      "loss": 0.0155,
+      "step": 2250
+    },
+    {
+      "epoch": 0.447114907140729,
+      "grad_norm": 0.992649917079707,
+      "learning_rate": 6.091532601546476e-06,
+      "loss": 0.0161,
+      "step": 2251
+    },
+    {
+      "epoch": 0.4473135365974774,
+      "grad_norm": 0.4457080027649029,
+      "learning_rate": 6.088392455416194e-06,
+      "loss": 0.0073,
+      "step": 2252
+    },
+    {
+      "epoch": 0.44751216605422584,
+      "grad_norm": 0.2552247040191699,
+      "learning_rate": 6.085251858583685e-06,
+      "loss": 0.0092,
+      "step": 2253
+    },
+    {
+      "epoch": 0.44771079551097426,
+      "grad_norm": 0.590972028978692,
+      "learning_rate": 6.082110812349468e-06,
+      "loss": 0.0109,
+      "step": 2254
+    },
+    {
+      "epoch": 0.44790942496772274,
+      "grad_norm": 0.4091054658923765,
+      "learning_rate": 6.078969318014246e-06,
+      "loss": 0.0124,
+      "step": 2255
+    },
+    {
+      "epoch": 0.44810805442447116,
+      "grad_norm": 0.7196903790599142,
+      "learning_rate": 6.075827376878911e-06,
+      "loss": 0.0148,
+      "step": 2256
+    },
+    {
+      "epoch": 0.4483066838812196,
+      "grad_norm": 0.3261047677288729,
+      "learning_rate": 6.072684990244537e-06,
+      "loss": 0.0098,
+      "step": 2257
+    },
+    {
+      "epoch": 0.448505313337968,
+      "grad_norm": 0.29300124021584395,
+      "learning_rate": 6.069542159412384e-06,
+      "loss": 0.0122,
+      "step": 2258
+    },
+    {
+      "epoch": 0.4487039427947165,
+      "grad_norm": 0.978139352013834,
+      "learning_rate": 6.066398885683892e-06,
+      "loss": 0.0144,
+      "step": 2259
+    },
+    {
+      "epoch": 0.4489025722514649,
+      "grad_norm": 0.7871382091207846,
+      "learning_rate": 6.06325517036069e-06,
+      "loss": 0.0155,
+      "step": 2260
+    },
+    {
+      "epoch": 0.4491012017082133,
+      "grad_norm": 1.1988153830297887,
+      "learning_rate": 6.060111014744586e-06,
+      "loss": 0.0114,
+      "step": 2261
+    },
+    {
+      "epoch": 0.44929983116496175,
+      "grad_norm": 0.5272884595442661,
+      "learning_rate": 6.056966420137572e-06,
+      "loss": 0.013,
+      "step": 2262
+    },
+    {
+      "epoch": 0.44949846062171017,
+      "grad_norm": 0.5853642484110945,
+      "learning_rate": 6.053821387841823e-06,
+      "loss": 0.0107,
+      "step": 2263
+    },
+    {
+      "epoch": 0.44969709007845865,
+      "grad_norm": 0.3163330948551929,
+      "learning_rate": 6.0506759191596944e-06,
+      "loss": 0.0122,
+      "step": 2264
+    },
+    {
+      "epoch": 0.4498957195352071,
+      "grad_norm": 0.43256342417833826,
+      "learning_rate": 6.047530015393718e-06,
+      "loss": 0.0081,
+      "step": 2265
+    },
+    {
+      "epoch": 0.4500943489919555,
+      "grad_norm": 0.8015470911760131,
+      "learning_rate": 6.04438367784661e-06,
+      "loss": 0.012,
+      "step": 2266
+    },
+    {
+      "epoch": 0.4502929784487039,
+      "grad_norm": 0.40936392070753425,
+      "learning_rate": 6.041236907821267e-06,
+      "loss": 0.0098,
+      "step": 2267
+    },
+    {
+      "epoch": 0.4504916079054524,
+      "grad_norm": 1.3184403290579898,
+      "learning_rate": 6.0380897066207654e-06,
+      "loss": 0.0159,
+      "step": 2268
+    },
+    {
+      "epoch": 0.4506902373622008,
+      "grad_norm": 0.36040823408450195,
+      "learning_rate": 6.034942075548355e-06,
+      "loss": 0.0063,
+      "step": 2269
+    },
+    {
+      "epoch": 0.45088886681894924,
+      "grad_norm": 0.5708270902664976,
+      "learning_rate": 6.0317940159074675e-06,
+      "loss": 0.0203,
+      "step": 2270
+    },
+    {
+      "epoch": 0.45108749627569766,
+      "grad_norm": 0.4185899585374011,
+      "learning_rate": 6.028645529001711e-06,
+      "loss": 0.009,
+      "step": 2271
+    },
+    {
+      "epoch": 0.45128612573244614,
+      "grad_norm": 0.5374794859076718,
+      "learning_rate": 6.02549661613487e-06,
+      "loss": 0.0184,
+      "step": 2272
+    },
+    {
+      "epoch": 0.45148475518919456,
+      "grad_norm": 0.3634143098642771,
+      "learning_rate": 6.0223472786109095e-06,
+      "loss": 0.0101,
+      "step": 2273
+    },
+    {
+      "epoch": 0.451683384645943,
+      "grad_norm": 0.42164752734379324,
+      "learning_rate": 6.019197517733963e-06,
+      "loss": 0.0124,
+      "step": 2274
+    },
+    {
+      "epoch": 0.4518820141026914,
+      "grad_norm": 0.2932256480659242,
+      "learning_rate": 6.016047334808345e-06,
+      "loss": 0.0056,
+      "step": 2275
+    },
+    {
+      "epoch": 0.4520806435594399,
+      "grad_norm": 0.950748673633275,
+      "learning_rate": 6.012896731138545e-06,
+      "loss": 0.0131,
+      "step": 2276
+    },
+    {
+      "epoch": 0.4522792730161883,
+      "grad_norm": 0.5735740533696118,
+      "learning_rate": 6.0097457080292206e-06,
+      "loss": 0.0158,
+      "step": 2277
+    },
+    {
+      "epoch": 0.45247790247293673,
+      "grad_norm": 0.6300357309402919,
+      "learning_rate": 6.006594266785212e-06,
+      "loss": 0.0198,
+      "step": 2278
+    },
+    {
+      "epoch": 0.45267653192968516,
+      "grad_norm": 0.45140542258277755,
+      "learning_rate": 6.003442408711524e-06,
+      "loss": 0.0103,
+      "step": 2279
+    },
+    {
+      "epoch": 0.45287516138643363,
+      "grad_norm": 0.31203226057353545,
+      "learning_rate": 6.000290135113338e-06,
+      "loss": 0.0072,
+      "step": 2280
+    },
+    {
+      "epoch": 0.45307379084318206,
+      "grad_norm": 0.5315768399241099,
+      "learning_rate": 5.997137447296011e-06,
+      "loss": 0.0189,
+      "step": 2281
+    },
+    {
+      "epoch": 0.4532724202999305,
+      "grad_norm": 0.7502568266180624,
+      "learning_rate": 5.993984346565065e-06,
+      "loss": 0.0168,
+      "step": 2282
+    },
+    {
+      "epoch": 0.4534710497566789,
+      "grad_norm": 0.7922773049664108,
+      "learning_rate": 5.990830834226197e-06,
+      "loss": 0.0197,
+      "step": 2283
+    },
+    {
+      "epoch": 0.4536696792134273,
+      "grad_norm": 0.5256093593559897,
+      "learning_rate": 5.987676911585273e-06,
+      "loss": 0.0118,
+      "step": 2284
+    },
+    {
+      "epoch": 0.4538683086701758,
+      "grad_norm": 0.4936069122490565,
+      "learning_rate": 5.984522579948329e-06,
+      "loss": 0.0119,
+      "step": 2285
+    },
+    {
+      "epoch": 0.4540669381269242,
+      "grad_norm": 0.6240742984791606,
+      "learning_rate": 5.981367840621569e-06,
+      "loss": 0.0103,
+      "step": 2286
+    },
+    {
+      "epoch": 0.45426556758367265,
+      "grad_norm": 0.4580728528848644,
+      "learning_rate": 5.97821269491137e-06,
+      "loss": 0.0089,
+      "step": 2287
+    },
+    {
+      "epoch": 0.45446419704042107,
+      "grad_norm": 1.0311027752364688,
+      "learning_rate": 5.975057144124274e-06,
+      "loss": 0.0109,
+      "step": 2288
+    },
+    {
+      "epoch": 0.45466282649716955,
+      "grad_norm": 0.47911929621431937,
+      "learning_rate": 5.971901189566991e-06,
+      "loss": 0.0127,
+      "step": 2289
+    },
+    {
+      "epoch": 0.45486145595391797,
+      "grad_norm": 0.31717175346501003,
+      "learning_rate": 5.968744832546398e-06,
+      "loss": 0.0131,
+      "step": 2290
+    },
+    {
+      "epoch": 0.4550600854106664,
+      "grad_norm": 0.5084576453823957,
+      "learning_rate": 5.965588074369541e-06,
+      "loss": 0.0108,
+      "step": 2291
+    },
+    {
+      "epoch": 0.4552587148674148,
+      "grad_norm": 0.4321949217258047,
+      "learning_rate": 5.962430916343627e-06,
+      "loss": 0.0073,
+      "step": 2292
+    },
+    {
+      "epoch": 0.4554573443241633,
+      "grad_norm": 0.3587917028460577,
+      "learning_rate": 5.959273359776037e-06,
+      "loss": 0.0095,
+      "step": 2293
+    },
+    {
+      "epoch": 0.4556559737809117,
+      "grad_norm": 0.39093795226899214,
+      "learning_rate": 5.956115405974308e-06,
+      "loss": 0.0067,
+      "step": 2294
+    },
+    {
+      "epoch": 0.45585460323766014,
+      "grad_norm": 1.022914745931775,
+      "learning_rate": 5.952957056246147e-06,
+      "loss": 0.0163,
+      "step": 2295
+    },
+    {
+      "epoch": 0.45605323269440856,
+      "grad_norm": 0.35063776749240805,
+      "learning_rate": 5.949798311899424e-06,
+      "loss": 0.0096,
+      "step": 2296
+    },
+    {
+      "epoch": 0.45625186215115704,
+      "grad_norm": 0.34523519354493915,
+      "learning_rate": 5.94663917424217e-06,
+      "loss": 0.0114,
+      "step": 2297
+    },
+    {
+      "epoch": 0.45645049160790546,
+      "grad_norm": 0.4356127664422994,
+      "learning_rate": 5.9434796445825835e-06,
+      "loss": 0.0114,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4566491210646539,
+      "grad_norm": 0.34697933488413446,
+      "learning_rate": 5.940319724229019e-06,
+      "loss": 0.007,
+      "step": 2299
+    },
+    {
+      "epoch": 0.4568477505214023,
+      "grad_norm": 0.45136277819236226,
+      "learning_rate": 5.937159414490001e-06,
+      "loss": 0.0141,
+      "step": 2300
+    },
+    {
+      "epoch": 0.4570463799781508,
+      "grad_norm": 0.9498700129253315,
+      "learning_rate": 5.933998716674206e-06,
+      "loss": 0.0113,
+      "step": 2301
+    },
+    {
+      "epoch": 0.4572450094348992,
+      "grad_norm": 1.0293691155517941,
+      "learning_rate": 5.930837632090479e-06,
+      "loss": 0.0221,
+      "step": 2302
+    },
+    {
+      "epoch": 0.45744363889164763,
+      "grad_norm": 0.3086672015952359,
+      "learning_rate": 5.92767616204782e-06,
+      "loss": 0.0081,
+      "step": 2303
+    },
+    {
+      "epoch": 0.45764226834839605,
+      "grad_norm": 0.3479833097855302,
+      "learning_rate": 5.92451430785539e-06,
+      "loss": 0.0062,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4578408978051445,
+      "grad_norm": 0.5660458407661002,
+      "learning_rate": 5.921352070822513e-06,
+      "loss": 0.0141,
+      "step": 2305
+    },
+    {
+      "epoch": 0.45803952726189295,
+      "grad_norm": 1.281389383626456,
+      "learning_rate": 5.918189452258665e-06,
+      "loss": 0.0192,
+      "step": 2306
+    },
+    {
+      "epoch": 0.4582381567186414,
+      "grad_norm": 0.6209042441203885,
+      "learning_rate": 5.915026453473485e-06,
+      "loss": 0.0237,
+      "step": 2307
+    },
+    {
+      "epoch": 0.4584367861753898,
+      "grad_norm": 0.372320922451502,
+      "learning_rate": 5.9118630757767666e-06,
+      "loss": 0.0078,
+      "step": 2308
+    },
+    {
+      "epoch": 0.4586354156321382,
+      "grad_norm": 0.5819030047417552,
+      "learning_rate": 5.908699320478462e-06,
+      "loss": 0.0073,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4588340450888867,
+      "grad_norm": 0.7061784530868636,
+      "learning_rate": 5.905535188888678e-06,
+      "loss": 0.0129,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4590326745456351,
+      "grad_norm": 0.6778701936405351,
+      "learning_rate": 5.9023706823176795e-06,
+      "loss": 0.0114,
+      "step": 2311
+    },
+    {
+      "epoch": 0.45923130400238354,
+      "grad_norm": 0.7264468288248946,
+      "learning_rate": 5.8992058020758855e-06,
+      "loss": 0.0231,
+      "step": 2312
+    },
+    {
+      "epoch": 0.45942993345913197,
+      "grad_norm": 0.6212393297209905,
+      "learning_rate": 5.89604054947387e-06,
+      "loss": 0.0143,
+      "step": 2313
+    },
+    {
+      "epoch": 0.45962856291588045,
+      "grad_norm": 0.35909749045150213,
+      "learning_rate": 5.892874925822359e-06,
+      "loss": 0.0122,
+      "step": 2314
+    },
+    {
+      "epoch": 0.45982719237262887,
+      "grad_norm": 0.3641939953347697,
+      "learning_rate": 5.889708932432235e-06,
+      "loss": 0.0105,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4600258218293773,
+      "grad_norm": 0.48230559364020364,
+      "learning_rate": 5.8865425706145355e-06,
+      "loss": 0.0127,
+      "step": 2316
+    },
+    {
+      "epoch": 0.4602244512861257,
+      "grad_norm": 0.46675212979529435,
+      "learning_rate": 5.883375841680442e-06,
+      "loss": 0.0124,
+      "step": 2317
+    },
+    {
+      "epoch": 0.4604230807428742,
+      "grad_norm": 0.3866132495591225,
+      "learning_rate": 5.880208746941299e-06,
+      "loss": 0.0073,
+      "step": 2318
+    },
+    {
+      "epoch": 0.4606217101996226,
+      "grad_norm": 0.439249606081369,
+      "learning_rate": 5.877041287708595e-06,
+      "loss": 0.0137,
+      "step": 2319
+    },
+    {
+      "epoch": 0.46082033965637104,
+      "grad_norm": 0.9131678817021807,
+      "learning_rate": 5.8738734652939725e-06,
+      "loss": 0.0144,
+      "step": 2320
+    },
+    {
+      "epoch": 0.46101896911311946,
+      "grad_norm": 0.4209332036129976,
+      "learning_rate": 5.870705281009222e-06,
+      "loss": 0.011,
+      "step": 2321
+    },
+    {
+      "epoch": 0.46121759856986794,
+      "grad_norm": 0.5289405995879706,
+      "learning_rate": 5.867536736166287e-06,
+      "loss": 0.0141,
+      "step": 2322
+    },
+    {
+      "epoch": 0.46141622802661636,
+      "grad_norm": 0.4232956656008544,
+      "learning_rate": 5.864367832077259e-06,
+      "loss": 0.0118,
+      "step": 2323
+    },
+    {
+      "epoch": 0.4616148574833648,
+      "grad_norm": 0.38126143858947364,
+      "learning_rate": 5.861198570054377e-06,
+      "loss": 0.0094,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4618134869401132,
+      "grad_norm": 0.5109437768251036,
+      "learning_rate": 5.858028951410029e-06,
+      "loss": 0.019,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4620121163968616,
+      "grad_norm": 0.9866712175295075,
+      "learning_rate": 5.854858977456753e-06,
+      "loss": 0.011,
+      "step": 2326
+    },
+    {
+      "epoch": 0.4622107458536101,
+      "grad_norm": 0.5509612624995485,
+      "learning_rate": 5.85168864950723e-06,
+      "loss": 0.0146,
+      "step": 2327
+    },
+    {
+      "epoch": 0.46240937531035853,
+      "grad_norm": 0.6697882679503572,
+      "learning_rate": 5.8485179688742896e-06,
+      "loss": 0.0156,
+      "step": 2328
+    },
+    {
+      "epoch": 0.46260800476710695,
+      "grad_norm": 0.25998329362845135,
+      "learning_rate": 5.845346936870907e-06,
+      "loss": 0.009,
+      "step": 2329
+    },
+    {
+      "epoch": 0.4628066342238554,
+      "grad_norm": 0.445245200827973,
+      "learning_rate": 5.8421755548102054e-06,
+      "loss": 0.0107,
+      "step": 2330
+    },
+    {
+      "epoch": 0.46300526368060385,
+      "grad_norm": 0.30406825765433,
+      "learning_rate": 5.8390038240054505e-06,
+      "loss": 0.0087,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4632038931373523,
+      "grad_norm": 0.5121143693890272,
+      "learning_rate": 5.835831745770052e-06,
+      "loss": 0.0121,
+      "step": 2332
+    },
+    {
+      "epoch": 0.4634025225941007,
+      "grad_norm": 0.6199846252981046,
+      "learning_rate": 5.832659321417565e-06,
+      "loss": 0.0112,
+      "step": 2333
+    },
+    {
+      "epoch": 0.4636011520508491,
+      "grad_norm": 0.48346062372856696,
+      "learning_rate": 5.829486552261685e-06,
+      "loss": 0.0131,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4637997815075976,
+      "grad_norm": 0.8757641479679567,
+      "learning_rate": 5.826313439616256e-06,
+      "loss": 0.0156,
+      "step": 2335
+    },
+    {
+      "epoch": 0.463998410964346,
+      "grad_norm": 0.43652114504307693,
+      "learning_rate": 5.8231399847952585e-06,
+      "loss": 0.0114,
+      "step": 2336
+    },
+    {
+      "epoch": 0.46419704042109444,
+      "grad_norm": 0.47839962377266626,
+      "learning_rate": 5.819966189112818e-06,
+      "loss": 0.0102,
+      "step": 2337
+    },
+    {
+      "epoch": 0.46439566987784286,
+      "grad_norm": 1.2127218820606134,
+      "learning_rate": 5.8167920538832e-06,
+      "loss": 0.014,
+      "step": 2338
+    },
+    {
+      "epoch": 0.46459429933459134,
+      "grad_norm": 0.769446434011527,
+      "learning_rate": 5.8136175804208106e-06,
+      "loss": 0.0114,
+      "step": 2339
+    },
+    {
+      "epoch": 0.46479292879133977,
+      "grad_norm": 0.34763955904579885,
+      "learning_rate": 5.8104427700401945e-06,
+      "loss": 0.015,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4649915582480882,
+      "grad_norm": 0.5597168255035329,
+      "learning_rate": 5.80726762405604e-06,
+      "loss": 0.014,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4651901877048366,
+      "grad_norm": 0.9135852247092504,
+      "learning_rate": 5.80409214378317e-06,
+      "loss": 0.0199,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4653888171615851,
+      "grad_norm": 0.6722348948820552,
+      "learning_rate": 5.800916330536549e-06,
+      "loss": 0.0111,
+      "step": 2343
+    },
+    {
+      "epoch": 0.4655874466183335,
+      "grad_norm": 0.2794802612129201,
+      "learning_rate": 5.7977401856312775e-06,
+      "loss": 0.0081,
+      "step": 2344
+    },
+    {
+      "epoch": 0.46578607607508193,
+      "grad_norm": 0.4778286582110137,
+      "learning_rate": 5.794563710382595e-06,
+      "loss": 0.0113,
+      "step": 2345
+    },
+    {
+      "epoch": 0.46598470553183036,
+      "grad_norm": 0.3799982420436121,
+      "learning_rate": 5.791386906105875e-06,
+      "loss": 0.0112,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4661833349885788,
+      "grad_norm": 0.5794990884883824,
+      "learning_rate": 5.788209774116629e-06,
+      "loss": 0.0167,
+      "step": 2347
+    },
+    {
+      "epoch": 0.46638196444532726,
+      "grad_norm": 0.9348715395945014,
+      "learning_rate": 5.785032315730506e-06,
+      "loss": 0.0248,
+      "step": 2348
+    },
+    {
+      "epoch": 0.4665805939020757,
+      "grad_norm": 0.5055416806585602,
+      "learning_rate": 5.781854532263287e-06,
+      "loss": 0.0158,
+      "step": 2349
+    },
+    {
+      "epoch": 0.4667792233588241,
+      "grad_norm": 0.6028354260098756,
+      "learning_rate": 5.778676425030888e-06,
+      "loss": 0.0103,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4669778528155725,
+      "grad_norm": 0.42355897766306966,
+      "learning_rate": 5.775497995349363e-06,
+      "loss": 0.0058,
+      "step": 2351
+    },
+    {
+      "epoch": 0.467176482272321,
+      "grad_norm": 0.30607023801120187,
+      "learning_rate": 5.772319244534893e-06,
+      "loss": 0.0116,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4673751117290694,
+      "grad_norm": 0.8992242787812346,
+      "learning_rate": 5.769140173903799e-06,
+      "loss": 0.0221,
+      "step": 2353
+    },
+    {
+      "epoch": 0.46757374118581785,
+      "grad_norm": 0.8268467494450803,
+      "learning_rate": 5.765960784772527e-06,
+      "loss": 0.01,
+      "step": 2354
+    },
+    {
+      "epoch": 0.46777237064256627,
+      "grad_norm": 0.31652768813051085,
+      "learning_rate": 5.762781078457662e-06,
+      "loss": 0.0071,
+      "step": 2355
+    },
+    {
+      "epoch": 0.46797100009931475,
+      "grad_norm": 0.5660464775694583,
+      "learning_rate": 5.759601056275916e-06,
+      "loss": 0.0099,
+      "step": 2356
+    },
+    {
+      "epoch": 0.46816962955606317,
+      "grad_norm": 0.6382863746802383,
+      "learning_rate": 5.756420719544135e-06,
+      "loss": 0.0161,
+      "step": 2357
+    },
+    {
+      "epoch": 0.4683682590128116,
+      "grad_norm": 0.6570237568395549,
+      "learning_rate": 5.75324006957929e-06,
+      "loss": 0.0151,
+      "step": 2358
+    },
+    {
+      "epoch": 0.46856688846956,
+      "grad_norm": 0.455513564680463,
+      "learning_rate": 5.7500591076984865e-06,
+      "loss": 0.0137,
+      "step": 2359
+    },
+    {
+      "epoch": 0.4687655179263085,
+      "grad_norm": 0.3782815896106589,
+      "learning_rate": 5.746877835218959e-06,
+      "loss": 0.0082,
+      "step": 2360
+    },
+    {
+      "epoch": 0.4689641473830569,
+      "grad_norm": 0.25011529623111944,
+      "learning_rate": 5.7436962534580665e-06,
+      "loss": 0.0099,
+      "step": 2361
+    },
+    {
+      "epoch": 0.46916277683980534,
+      "grad_norm": 0.2755162028039524,
+      "learning_rate": 5.7405143637333e-06,
+      "loss": 0.0074,
+      "step": 2362
+    },
+    {
+      "epoch": 0.46936140629655376,
+      "grad_norm": 0.4887019159723483,
+      "learning_rate": 5.737332167362278e-06,
+      "loss": 0.0123,
+      "step": 2363
+    },
+    {
+      "epoch": 0.46956003575330224,
+      "grad_norm": 0.6804047453581535,
+      "learning_rate": 5.734149665662744e-06,
+      "loss": 0.0099,
+      "step": 2364
+    },
+    {
+      "epoch": 0.46975866521005066,
+      "grad_norm": 0.8398249561968816,
+      "learning_rate": 5.730966859952568e-06,
+      "loss": 0.0193,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4699572946667991,
+      "grad_norm": 0.30808476371188004,
+      "learning_rate": 5.727783751549748e-06,
+      "loss": 0.0101,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4701559241235475,
+      "grad_norm": 0.40427106496463305,
+      "learning_rate": 5.724600341772404e-06,
+      "loss": 0.0072,
+      "step": 2367
+    },
+    {
+      "epoch": 0.47035455358029593,
+      "grad_norm": 0.5655774899898666,
+      "learning_rate": 5.721416631938785e-06,
+      "loss": 0.0161,
+      "step": 2368
+    },
+    {
+      "epoch": 0.4705531830370444,
+      "grad_norm": 0.6325786828061096,
+      "learning_rate": 5.718232623367262e-06,
+      "loss": 0.0183,
+      "step": 2369
+    },
+    {
+      "epoch": 0.47075181249379283,
+      "grad_norm": 0.5808533874064912,
+      "learning_rate": 5.715048317376327e-06,
+      "loss": 0.0074,
+      "step": 2370
+    },
+    {
+      "epoch": 0.47095044195054125,
+      "grad_norm": 0.8445148294332612,
+      "learning_rate": 5.7118637152846015e-06,
+      "loss": 0.0161,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4711490714072897,
+      "grad_norm": 0.4846773932797683,
+      "learning_rate": 5.708678818410823e-06,
+      "loss": 0.0057,
+      "step": 2372
+    },
+    {
+      "epoch": 0.47134770086403815,
+      "grad_norm": 0.4942960720168355,
+      "learning_rate": 5.705493628073856e-06,
+      "loss": 0.0163,
+      "step": 2373
+    },
+    {
+      "epoch": 0.4715463303207866,
+      "grad_norm": 0.2926160774657436,
+      "learning_rate": 5.702308145592684e-06,
+      "loss": 0.0084,
+      "step": 2374
+    },
+    {
+      "epoch": 0.471744959777535,
+      "grad_norm": 0.36092426480142087,
+      "learning_rate": 5.699122372286413e-06,
+      "loss": 0.0156,
+      "step": 2375
+    },
+    {
+      "epoch": 0.4719435892342834,
+      "grad_norm": 0.971391873556755,
+      "learning_rate": 5.6959363094742684e-06,
+      "loss": 0.0209,
+      "step": 2376
+    },
+    {
+      "epoch": 0.4721422186910319,
+      "grad_norm": 0.5192492237273335,
+      "learning_rate": 5.692749958475593e-06,
+      "loss": 0.0158,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4723408481477803,
+      "grad_norm": 0.37516394436059797,
+      "learning_rate": 5.689563320609854e-06,
+      "loss": 0.012,
+      "step": 2378
+    },
+    {
+      "epoch": 0.47253947760452875,
+      "grad_norm": 0.7321340081454777,
+      "learning_rate": 5.686376397196635e-06,
+      "loss": 0.0095,
+      "step": 2379
+    },
+    {
+      "epoch": 0.47273810706127717,
+      "grad_norm": 0.6380956047969817,
+      "learning_rate": 5.683189189555637e-06,
+      "loss": 0.0094,
+      "step": 2380
+    },
+    {
+      "epoch": 0.47293673651802565,
+      "grad_norm": 0.5714109992216438,
+      "learning_rate": 5.68000169900668e-06,
+      "loss": 0.008,
+      "step": 2381
+    },
+    {
+      "epoch": 0.47313536597477407,
+      "grad_norm": 0.773380440894765,
+      "learning_rate": 5.6768139268697e-06,
+      "loss": 0.0152,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4733339954315225,
+      "grad_norm": 0.6960015446023513,
+      "learning_rate": 5.673625874464751e-06,
+      "loss": 0.0217,
+      "step": 2383
+    },
+    {
+      "epoch": 0.4735326248882709,
+      "grad_norm": 0.6037591137291943,
+      "learning_rate": 5.670437543112003e-06,
+      "loss": 0.0136,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4737312543450194,
+      "grad_norm": 0.541539265791468,
+      "learning_rate": 5.667248934131739e-06,
+      "loss": 0.0087,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4739298838017678,
+      "grad_norm": 0.4801580538978032,
+      "learning_rate": 5.6640600488443585e-06,
+      "loss": 0.0106,
+      "step": 2386
+    },
+    {
+      "epoch": 0.47412851325851624,
+      "grad_norm": 0.4106221925173567,
+      "learning_rate": 5.660870888570378e-06,
+      "loss": 0.0121,
+      "step": 2387
+    },
+    {
+      "epoch": 0.47432714271526466,
+      "grad_norm": 1.2008332110233264,
+      "learning_rate": 5.657681454630424e-06,
+      "loss": 0.0164,
+      "step": 2388
+    },
+    {
+      "epoch": 0.4745257721720131,
+      "grad_norm": 0.46847613164327123,
+      "learning_rate": 5.654491748345238e-06,
+      "loss": 0.0112,
+      "step": 2389
+    },
+    {
+      "epoch": 0.47472440162876156,
+      "grad_norm": 0.4866905116366235,
+      "learning_rate": 5.651301771035675e-06,
+      "loss": 0.0123,
+      "step": 2390
+    },
+    {
+      "epoch": 0.47492303108551,
+      "grad_norm": 0.8229938698433211,
+      "learning_rate": 5.6481115240227005e-06,
+      "loss": 0.0144,
+      "step": 2391
+    },
+    {
+      "epoch": 0.4751216605422584,
+      "grad_norm": 0.9137037442295208,
+      "learning_rate": 5.644921008627391e-06,
+      "loss": 0.0151,
+      "step": 2392
+    },
+    {
+      "epoch": 0.47532028999900683,
+      "grad_norm": 0.4876172441311117,
+      "learning_rate": 5.6417302261709404e-06,
+      "loss": 0.0095,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4755189194557553,
+      "grad_norm": 0.37524536677463444,
+      "learning_rate": 5.638539177974645e-06,
+      "loss": 0.0121,
+      "step": 2394
+    },
+    {
+      "epoch": 0.47571754891250373,
+      "grad_norm": 0.7581160723528595,
+      "learning_rate": 5.635347865359915e-06,
+      "loss": 0.0107,
+      "step": 2395
+    },
+    {
+      "epoch": 0.47591617836925215,
+      "grad_norm": 0.4697124876851263,
+      "learning_rate": 5.632156289648272e-06,
+      "loss": 0.0141,
+      "step": 2396
+    },
+    {
+      "epoch": 0.4761148078260006,
+      "grad_norm": 1.2004601762194573,
+      "learning_rate": 5.62896445216134e-06,
+      "loss": 0.0178,
+      "step": 2397
+    },
+    {
+      "epoch": 0.47631343728274905,
+      "grad_norm": 0.44792347789247144,
+      "learning_rate": 5.6257723542208595e-06,
+      "loss": 0.0094,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4765120667394975,
+      "grad_norm": 0.6253105011573149,
+      "learning_rate": 5.622579997148674e-06,
+      "loss": 0.0104,
+      "step": 2399
+    },
+    {
+      "epoch": 0.4767106961962459,
+      "grad_norm": 0.38725763342264224,
+      "learning_rate": 5.619387382266734e-06,
+      "loss": 0.0181,
+      "step": 2400
+    },
+    {
+      "epoch": 0.4769093256529943,
+      "grad_norm": 0.5602962469904611,
+      "learning_rate": 5.6161945108971005e-06,
+      "loss": 0.0195,
+      "step": 2401
+    },
+    {
+      "epoch": 0.4771079551097428,
+      "grad_norm": 0.48104235373399207,
+      "learning_rate": 5.6130013843619366e-06,
+      "loss": 0.0197,
+      "step": 2402
+    },
+    {
+      "epoch": 0.4773065845664912,
+      "grad_norm": 0.3039832053300856,
+      "learning_rate": 5.609808003983513e-06,
+      "loss": 0.0067,
+      "step": 2403
+    },
+    {
+      "epoch": 0.47750521402323964,
+      "grad_norm": 0.7235246350240075,
+      "learning_rate": 5.606614371084206e-06,
+      "loss": 0.0193,
+      "step": 2404
+    },
+    {
+      "epoch": 0.47770384347998807,
+      "grad_norm": 0.547167273803981,
+      "learning_rate": 5.603420486986495e-06,
+      "loss": 0.0059,
+      "step": 2405
+    },
+    {
+      "epoch": 0.47790247293673654,
+      "grad_norm": 0.5020367397204322,
+      "learning_rate": 5.600226353012965e-06,
+      "loss": 0.0124,
+      "step": 2406
+    },
+    {
+      "epoch": 0.47810110239348497,
+      "grad_norm": 0.29441250473742775,
+      "learning_rate": 5.597031970486304e-06,
+      "loss": 0.0089,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4782997318502334,
+      "grad_norm": 0.3051620947194149,
+      "learning_rate": 5.593837340729302e-06,
+      "loss": 0.0102,
+      "step": 2408
+    },
+    {
+      "epoch": 0.4784983613069818,
+      "grad_norm": 0.46639149386889667,
+      "learning_rate": 5.590642465064852e-06,
+      "loss": 0.0154,
+      "step": 2409
+    },
+    {
+      "epoch": 0.47869699076373023,
+      "grad_norm": 0.6356240089361235,
+      "learning_rate": 5.587447344815946e-06,
+      "loss": 0.0158,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4788956202204787,
+      "grad_norm": 0.7929409757443711,
+      "learning_rate": 5.584251981305685e-06,
+      "loss": 0.016,
+      "step": 2411
+    },
+    {
+      "epoch": 0.47909424967722714,
+      "grad_norm": 0.7886868857089763,
+      "learning_rate": 5.581056375857263e-06,
+      "loss": 0.0158,
+      "step": 2412
+    },
+    {
+      "epoch": 0.47929287913397556,
+      "grad_norm": 0.6545632498798621,
+      "learning_rate": 5.577860529793978e-06,
+      "loss": 0.0142,
+      "step": 2413
+    },
+    {
+      "epoch": 0.479491508590724,
+      "grad_norm": 0.6445078069693473,
+      "learning_rate": 5.574664444439226e-06,
+      "loss": 0.0128,
+      "step": 2414
+    },
+    {
+      "epoch": 0.47969013804747246,
+      "grad_norm": 0.4114117395918578,
+      "learning_rate": 5.571468121116504e-06,
+      "loss": 0.0118,
+      "step": 2415
+    },
+    {
+      "epoch": 0.4798887675042209,
+      "grad_norm": 0.49308068014298745,
+      "learning_rate": 5.568271561149404e-06,
+      "loss": 0.0199,
+      "step": 2416
+    },
+    {
+      "epoch": 0.4800873969609693,
+      "grad_norm": 0.8300192404106222,
+      "learning_rate": 5.56507476586162e-06,
+      "loss": 0.0163,
+      "step": 2417
+    },
+    {
+      "epoch": 0.4802860264177177,
+      "grad_norm": 0.7203031117955251,
+      "learning_rate": 5.561877736576942e-06,
+      "loss": 0.0192,
+      "step": 2418
+    },
+    {
+      "epoch": 0.4804846558744662,
+      "grad_norm": 0.63747435264011,
+      "learning_rate": 5.558680474619255e-06,
+      "loss": 0.0116,
+      "step": 2419
+    },
+    {
+      "epoch": 0.4806832853312146,
+      "grad_norm": 0.6686966563642719,
+      "learning_rate": 5.5554829813125446e-06,
+      "loss": 0.0137,
+      "step": 2420
+    },
+    {
+      "epoch": 0.48088191478796305,
+      "grad_norm": 0.7236264839110779,
+      "learning_rate": 5.552285257980887e-06,
+      "loss": 0.0118,
+      "step": 2421
+    },
+    {
+      "epoch": 0.48108054424471147,
+      "grad_norm": 0.3489337234437465,
+      "learning_rate": 5.549087305948455e-06,
+      "loss": 0.0125,
+      "step": 2422
+    },
+    {
+      "epoch": 0.48127917370145995,
+      "grad_norm": 0.6402623274118869,
+      "learning_rate": 5.545889126539522e-06,
+      "loss": 0.0158,
+      "step": 2423
+    },
+    {
+      "epoch": 0.4814778031582084,
+      "grad_norm": 0.6755621454969479,
+      "learning_rate": 5.542690721078447e-06,
+      "loss": 0.018,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4816764326149568,
+      "grad_norm": 0.3813390848817102,
+      "learning_rate": 5.5394920908896895e-06,
+      "loss": 0.0082,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4818750620717052,
+      "grad_norm": 0.5944323310591813,
+      "learning_rate": 5.536293237297796e-06,
+      "loss": 0.0086,
+      "step": 2426
+    },
+    {
+      "epoch": 0.4820736915284537,
+      "grad_norm": 0.5823022738844985,
+      "learning_rate": 5.533094161627412e-06,
+      "loss": 0.0129,
+      "step": 2427
+    },
+    {
+      "epoch": 0.4822723209852021,
+      "grad_norm": 0.24109484982366644,
+      "learning_rate": 5.529894865203267e-06,
+      "loss": 0.0059,
+      "step": 2428
+    },
+    {
+      "epoch": 0.48247095044195054,
+      "grad_norm": 0.6279027842239673,
+      "learning_rate": 5.52669534935019e-06,
+      "loss": 0.0151,
+      "step": 2429
+    },
+    {
+      "epoch": 0.48266957989869896,
+      "grad_norm": 0.8846786775845136,
+      "learning_rate": 5.523495615393095e-06,
+      "loss": 0.0208,
+      "step": 2430
+    },
+    {
+      "epoch": 0.4828682093554474,
+      "grad_norm": 0.38091540228676024,
+      "learning_rate": 5.520295664656992e-06,
+      "loss": 0.0069,
+      "step": 2431
+    },
+    {
+      "epoch": 0.48306683881219586,
+      "grad_norm": 0.30556096556441176,
+      "learning_rate": 5.517095498466976e-06,
+      "loss": 0.0061,
+      "step": 2432
+    },
+    {
+      "epoch": 0.4832654682689443,
+      "grad_norm": 0.8509282833418818,
+      "learning_rate": 5.51389511814823e-06,
+      "loss": 0.0135,
+      "step": 2433
+    },
+    {
+      "epoch": 0.4834640977256927,
+      "grad_norm": 0.37578690343483734,
+      "learning_rate": 5.51069452502603e-06,
+      "loss": 0.0091,
+      "step": 2434
+    },
+    {
+      "epoch": 0.48366272718244113,
+      "grad_norm": 0.46966875266445063,
+      "learning_rate": 5.5074937204257385e-06,
+      "loss": 0.0147,
+      "step": 2435
+    },
+    {
+      "epoch": 0.4838613566391896,
+      "grad_norm": 0.509057592751693,
+      "learning_rate": 5.504292705672807e-06,
+      "loss": 0.0142,
+      "step": 2436
+    },
+    {
+      "epoch": 0.48405998609593803,
+      "grad_norm": 0.472350938168894,
+      "learning_rate": 5.5010914820927695e-06,
+      "loss": 0.01,
+      "step": 2437
+    },
+    {
+      "epoch": 0.48425861555268646,
+      "grad_norm": 0.4293231269108412,
+      "learning_rate": 5.497890051011252e-06,
+      "loss": 0.009,
+      "step": 2438
+    },
+    {
+      "epoch": 0.4844572450094349,
+      "grad_norm": 0.6277901027183694,
+      "learning_rate": 5.494688413753964e-06,
+      "loss": 0.0085,
+      "step": 2439
+    },
+    {
+      "epoch": 0.48465587446618336,
+      "grad_norm": 0.4248433246133367,
+      "learning_rate": 5.491486571646698e-06,
+      "loss": 0.0103,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4848545039229318,
+      "grad_norm": 0.7745270788422335,
+      "learning_rate": 5.488284526015335e-06,
+      "loss": 0.0169,
+      "step": 2441
+    },
+    {
+      "epoch": 0.4850531333796802,
+      "grad_norm": 0.9595839872373445,
+      "learning_rate": 5.485082278185838e-06,
+      "loss": 0.0123,
+      "step": 2442
+    },
+    {
+      "epoch": 0.4852517628364286,
+      "grad_norm": 0.24937919516680546,
+      "learning_rate": 5.481879829484256e-06,
+      "loss": 0.0072,
+      "step": 2443
+    },
+    {
+      "epoch": 0.4854503922931771,
+      "grad_norm": 0.5052250608119354,
+      "learning_rate": 5.47867718123672e-06,
+      "loss": 0.0126,
+      "step": 2444
+    },
+    {
+      "epoch": 0.4856490217499255,
+      "grad_norm": 0.6095834648470921,
+      "learning_rate": 5.475474334769443e-06,
+      "loss": 0.0127,
+      "step": 2445
+    },
+    {
+      "epoch": 0.48584765120667395,
+      "grad_norm": 0.5239132674664736,
+      "learning_rate": 5.47227129140872e-06,
+      "loss": 0.0109,
+      "step": 2446
+    },
+    {
+      "epoch": 0.48604628066342237,
+      "grad_norm": 0.40049557705857963,
+      "learning_rate": 5.4690680524809246e-06,
+      "loss": 0.0072,
+      "step": 2447
+    },
+    {
+      "epoch": 0.48624491012017085,
+      "grad_norm": 0.6583755900248895,
+      "learning_rate": 5.465864619312522e-06,
+      "loss": 0.0112,
+      "step": 2448
+    },
+    {
+      "epoch": 0.48644353957691927,
+      "grad_norm": 0.511543136523891,
+      "learning_rate": 5.462660993230045e-06,
+      "loss": 0.0111,
+      "step": 2449
+    },
+    {
+      "epoch": 0.4866421690336677,
+      "grad_norm": 0.6333349669994685,
+      "learning_rate": 5.459457175560117e-06,
+      "loss": 0.0142,
+      "step": 2450
+    },
+    {
+      "epoch": 0.4868407984904161,
+      "grad_norm": 0.4081955737456362,
+      "learning_rate": 5.456253167629431e-06,
+      "loss": 0.0106,
+      "step": 2451
+    },
+    {
+      "epoch": 0.48703942794716454,
+      "grad_norm": 0.45569800327311816,
+      "learning_rate": 5.4530489707647646e-06,
+      "loss": 0.0122,
+      "step": 2452
+    },
+    {
+      "epoch": 0.487238057403913,
+      "grad_norm": 0.49468712119644775,
+      "learning_rate": 5.449844586292974e-06,
+      "loss": 0.0163,
+      "step": 2453
+    },
+    {
+      "epoch": 0.48743668686066144,
+      "grad_norm": 1.3504172026817023,
+      "learning_rate": 5.446640015540989e-06,
+      "loss": 0.0165,
+      "step": 2454
+    },
+    {
+      "epoch": 0.48763531631740986,
+      "grad_norm": 0.48452241295226994,
+      "learning_rate": 5.443435259835822e-06,
+      "loss": 0.0116,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4878339457741583,
+      "grad_norm": 0.5427977276769387,
+      "learning_rate": 5.44023032050456e-06,
+      "loss": 0.0129,
+      "step": 2456
+    },
+    {
+      "epoch": 0.48803257523090676,
+      "grad_norm": 0.3796715316602044,
+      "learning_rate": 5.437025198874361e-06,
+      "loss": 0.0107,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4882312046876552,
+      "grad_norm": 0.9111058024326351,
+      "learning_rate": 5.433819896272464e-06,
+      "loss": 0.0157,
+      "step": 2458
+    },
+    {
+      "epoch": 0.4884298341444036,
+      "grad_norm": 0.490889205825391,
+      "learning_rate": 5.4306144140261845e-06,
+      "loss": 0.009,
+      "step": 2459
+    },
+    {
+      "epoch": 0.48862846360115203,
+      "grad_norm": 0.6591819950885145,
+      "learning_rate": 5.427408753462905e-06,
+      "loss": 0.0151,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4888270930579005,
+      "grad_norm": 0.3574688903762408,
+      "learning_rate": 5.42420291591009e-06,
+      "loss": 0.0145,
+      "step": 2461
+    },
+    {
+      "epoch": 0.48902572251464893,
+      "grad_norm": 0.7539287197857844,
+      "learning_rate": 5.420996902695273e-06,
+      "loss": 0.0134,
+      "step": 2462
+    },
+    {
+      "epoch": 0.48922435197139735,
+      "grad_norm": 0.6981359145878261,
+      "learning_rate": 5.4177907151460585e-06,
+      "loss": 0.0119,
+      "step": 2463
+    },
+    {
+      "epoch": 0.4894229814281458,
+      "grad_norm": 0.606244215044152,
+      "learning_rate": 5.414584354590129e-06,
+      "loss": 0.0111,
+      "step": 2464
+    },
+    {
+      "epoch": 0.48962161088489425,
+      "grad_norm": 0.5459437232022195,
+      "learning_rate": 5.411377822355233e-06,
+      "loss": 0.0091,
+      "step": 2465
+    },
+    {
+      "epoch": 0.4898202403416427,
+      "grad_norm": 0.5279168001785495,
+      "learning_rate": 5.408171119769192e-06,
+      "loss": 0.0112,
+      "step": 2466
+    },
+    {
+      "epoch": 0.4900188697983911,
+      "grad_norm": 0.4695727703267886,
+      "learning_rate": 5.4049642481598985e-06,
+      "loss": 0.0117,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4902174992551395,
+      "grad_norm": 0.43550533403024305,
+      "learning_rate": 5.401757208855317e-06,
+      "loss": 0.0048,
+      "step": 2468
+    },
+    {
+      "epoch": 0.490416128711888,
+      "grad_norm": 0.42677326587153497,
+      "learning_rate": 5.398550003183478e-06,
+      "loss": 0.0081,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4906147581686364,
+      "grad_norm": 0.6348135613049047,
+      "learning_rate": 5.39534263247248e-06,
+      "loss": 0.0209,
+      "step": 2470
+    },
+    {
+      "epoch": 0.49081338762538484,
+      "grad_norm": 0.5524103691775981,
+      "learning_rate": 5.392135098050495e-06,
+      "loss": 0.0119,
+      "step": 2471
+    },
+    {
+      "epoch": 0.49101201708213327,
+      "grad_norm": 0.7108743625403123,
+      "learning_rate": 5.388927401245757e-06,
+      "loss": 0.0176,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4912106465388817,
+      "grad_norm": 0.4353791961565994,
+      "learning_rate": 5.385719543386574e-06,
+      "loss": 0.0106,
+      "step": 2473
+    },
+    {
+      "epoch": 0.49140927599563017,
+      "grad_norm": 0.6306640021625024,
+      "learning_rate": 5.3825115258013126e-06,
+      "loss": 0.0103,
+      "step": 2474
+    },
+    {
+      "epoch": 0.4916079054523786,
+      "grad_norm": 0.5210772415889535,
+      "learning_rate": 5.379303349818412e-06,
+      "loss": 0.009,
+      "step": 2475
+    },
+    {
+      "epoch": 0.491806534909127,
+      "grad_norm": 0.414410686373316,
+      "learning_rate": 5.376095016766374e-06,
+      "loss": 0.0089,
+      "step": 2476
+    },
+    {
+      "epoch": 0.49200516436587544,
+      "grad_norm": 0.4551862364482468,
+      "learning_rate": 5.372886527973767e-06,
+      "loss": 0.0082,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4922037938226239,
+      "grad_norm": 0.7775904217757333,
+      "learning_rate": 5.369677884769221e-06,
+      "loss": 0.0177,
+      "step": 2478
+    },
+    {
+      "epoch": 0.49240242327937234,
+      "grad_norm": 1.0123752948840063,
+      "learning_rate": 5.366469088481433e-06,
+      "loss": 0.0137,
+      "step": 2479
+    },
+    {
+      "epoch": 0.49260105273612076,
+      "grad_norm": 0.4494991849679043,
+      "learning_rate": 5.3632601404391635e-06,
+      "loss": 0.0094,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4927996821928692,
+      "grad_norm": 0.7550282831087276,
+      "learning_rate": 5.360051041971234e-06,
+      "loss": 0.014,
+      "step": 2481
+    },
+    {
+      "epoch": 0.49299831164961766,
+      "grad_norm": 0.46682384084711037,
+      "learning_rate": 5.356841794406527e-06,
+      "loss": 0.0151,
+      "step": 2482
+    },
+    {
+      "epoch": 0.4931969411063661,
+      "grad_norm": 0.4115591498845582,
+      "learning_rate": 5.353632399073991e-06,
+      "loss": 0.0113,
+      "step": 2483
+    },
+    {
+      "epoch": 0.4933955705631145,
+      "grad_norm": 0.7247787357221698,
+      "learning_rate": 5.350422857302633e-06,
+      "loss": 0.0134,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4935942000198629,
+      "grad_norm": 0.8578245479997127,
+      "learning_rate": 5.347213170421519e-06,
+      "loss": 0.0148,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4937928294766114,
+      "grad_norm": 0.4699237127768561,
+      "learning_rate": 5.344003339759781e-06,
+      "loss": 0.0097,
+      "step": 2486
+    },
+    {
+      "epoch": 0.49399145893335983,
+      "grad_norm": 0.4744010503049874,
+      "learning_rate": 5.340793366646604e-06,
+      "loss": 0.0115,
+      "step": 2487
+    },
+    {
+      "epoch": 0.49419008839010825,
+      "grad_norm": 0.636008252077129,
+      "learning_rate": 5.337583252411235e-06,
+      "loss": 0.0102,
+      "step": 2488
+    },
+    {
+      "epoch": 0.4943887178468567,
+      "grad_norm": 0.47474845684362593,
+      "learning_rate": 5.33437299838298e-06,
+      "loss": 0.0124,
+      "step": 2489
+    },
+    {
+      "epoch": 0.49458734730360515,
+      "grad_norm": 1.0566024501358895,
+      "learning_rate": 5.3311626058911994e-06,
+      "loss": 0.0151,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4947859767603536,
+      "grad_norm": 0.36678370138916344,
+      "learning_rate": 5.327952076265317e-06,
+      "loss": 0.0147,
+      "step": 2491
+    },
+    {
+      "epoch": 0.494984606217102,
+      "grad_norm": 0.25272585797585356,
+      "learning_rate": 5.324741410834807e-06,
+      "loss": 0.0067,
+      "step": 2492
+    },
+    {
+      "epoch": 0.4951832356738504,
+      "grad_norm": 0.6048164575103134,
+      "learning_rate": 5.321530610929204e-06,
+      "loss": 0.0113,
+      "step": 2493
+    },
+    {
+      "epoch": 0.49538186513059884,
+      "grad_norm": 0.7695263702174417,
+      "learning_rate": 5.318319677878098e-06,
+      "loss": 0.0142,
+      "step": 2494
+    },
+    {
+      "epoch": 0.4955804945873473,
+      "grad_norm": 0.33525412117269704,
+      "learning_rate": 5.315108613011132e-06,
+      "loss": 0.0124,
+      "step": 2495
+    },
+    {
+      "epoch": 0.49577912404409574,
+      "grad_norm": 0.7878389738793319,
+      "learning_rate": 5.311897417658005e-06,
+      "loss": 0.0151,
+      "step": 2496
+    },
+    {
+      "epoch": 0.49597775350084417,
+      "grad_norm": 0.4578947485727716,
+      "learning_rate": 5.308686093148467e-06,
+      "loss": 0.0115,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4961763829575926,
+      "grad_norm": 1.030479841136134,
+      "learning_rate": 5.305474640812331e-06,
+      "loss": 0.0139,
+      "step": 2498
+    },
+    {
+      "epoch": 0.49637501241434107,
+      "grad_norm": 0.796961987633012,
+      "learning_rate": 5.30226306197945e-06,
+      "loss": 0.0109,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4965736418710895,
+      "grad_norm": 0.782382140972235,
+      "learning_rate": 5.299051357979738e-06,
+      "loss": 0.0187,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4967722713278379,
+      "grad_norm": 1.121114404342789,
+      "learning_rate": 5.295839530143159e-06,
+      "loss": 0.0178,
+      "step": 2501
+    },
+    {
+      "epoch": 0.49697090078458633,
+      "grad_norm": 0.7261947822033853,
+      "learning_rate": 5.2926275797997255e-06,
+      "loss": 0.0188,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4971695302413348,
+      "grad_norm": 0.46923022153419347,
+      "learning_rate": 5.289415508279505e-06,
+      "loss": 0.0073,
+      "step": 2503
+    },
+    {
+      "epoch": 0.49736815969808323,
+      "grad_norm": 0.7182836113385063,
+      "learning_rate": 5.286203316912613e-06,
+      "loss": 0.015,
+      "step": 2504
+    },
+    {
+      "epoch": 0.49756678915483166,
+      "grad_norm": 0.603411625751702,
+      "learning_rate": 5.282991007029213e-06,
+      "loss": 0.0106,
+      "step": 2505
+    },
+    {
+      "epoch": 0.4977654186115801,
+      "grad_norm": 0.8174796617956737,
+      "learning_rate": 5.27977857995952e-06,
+      "loss": 0.0149,
+      "step": 2506
+    },
+    {
+      "epoch": 0.49796404806832856,
+      "grad_norm": 1.0146932905608117,
+      "learning_rate": 5.276566037033798e-06,
+      "loss": 0.0141,
+      "step": 2507
+    },
+    {
+      "epoch": 0.498162677525077,
+      "grad_norm": 0.4163900253441471,
+      "learning_rate": 5.273353379582357e-06,
+      "loss": 0.0108,
+      "step": 2508
+    },
+    {
+      "epoch": 0.4983613069818254,
+      "grad_norm": 0.763281658551113,
+      "learning_rate": 5.270140608935555e-06,
+      "loss": 0.015,
+      "step": 2509
+    },
+    {
+      "epoch": 0.4985599364385738,
+      "grad_norm": 0.22259719035648037,
+      "learning_rate": 5.266927726423797e-06,
+      "loss": 0.0087,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4987585658953223,
+      "grad_norm": 0.42395383851091467,
+      "learning_rate": 5.263714733377535e-06,
+      "loss": 0.0133,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4989571953520707,
+      "grad_norm": 0.5328797312516291,
+      "learning_rate": 5.260501631127266e-06,
+      "loss": 0.012,
+      "step": 2512
+    },
+    {
+      "epoch": 0.49915582480881915,
+      "grad_norm": 0.4743705135796904,
+      "learning_rate": 5.257288421003534e-06,
+      "loss": 0.0196,
+      "step": 2513
+    },
+    {
+      "epoch": 0.49935445426556757,
+      "grad_norm": 0.5093401512061038,
+      "learning_rate": 5.254075104336922e-06,
+      "loss": 0.0082,
+      "step": 2514
+    },
+    {
+      "epoch": 0.499553083722316,
+      "grad_norm": 0.9029692999709029,
+      "learning_rate": 5.250861682458066e-06,
+      "loss": 0.0249,
+      "step": 2515
+    },
+    {
+      "epoch": 0.49975171317906447,
+      "grad_norm": 1.777870716049857,
+      "learning_rate": 5.247648156697637e-06,
+      "loss": 0.0258,
+      "step": 2516
+    },
+    {
+      "epoch": 0.4999503426358129,
+      "grad_norm": 0.5892960441699706,
+      "learning_rate": 5.2444345283863555e-06,
+      "loss": 0.008,
+      "step": 2517
+    },
+    {
+      "epoch": 0.5001489720925614,
+      "grad_norm": 0.44368783450479404,
+      "learning_rate": 5.241220798854979e-06,
+      "loss": 0.0097,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5003476015493098,
+      "grad_norm": 0.4921667565769311,
+      "learning_rate": 5.238006969434313e-06,
+      "loss": 0.0111,
+      "step": 2519
+    },
+    {
+      "epoch": 0.5005462310060582,
+      "grad_norm": 0.5508976317775016,
+      "learning_rate": 5.234793041455199e-06,
+      "loss": 0.0143,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5007448604628066,
+      "grad_norm": 0.7142397409065645,
+      "learning_rate": 5.2315790162485204e-06,
+      "loss": 0.0141,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5009434899195551,
+      "grad_norm": 0.47566862647539554,
+      "learning_rate": 5.228364895145203e-06,
+      "loss": 0.0162,
+      "step": 2522
+    },
+    {
+      "epoch": 0.5011421193763035,
+      "grad_norm": 0.9705471048491111,
+      "learning_rate": 5.225150679476212e-06,
+      "loss": 0.0165,
+      "step": 2523
+    },
+    {
+      "epoch": 0.5013407488330519,
+      "grad_norm": 0.5751231880679607,
+      "learning_rate": 5.22193637057255e-06,
+      "loss": 0.0138,
+      "step": 2524
+    },
+    {
+      "epoch": 0.5015393782898003,
+      "grad_norm": 0.40697823166729785,
+      "learning_rate": 5.218721969765259e-06,
+      "loss": 0.0055,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5017380077465489,
+      "grad_norm": 0.4094426383199233,
+      "learning_rate": 5.21550747838542e-06,
+      "loss": 0.0122,
+      "step": 2526
+    },
+    {
+      "epoch": 0.5019366372032973,
+      "grad_norm": 0.4031192231606706,
+      "learning_rate": 5.21229289776415e-06,
+      "loss": 0.0092,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5021352666600457,
+      "grad_norm": 0.35911635894078986,
+      "learning_rate": 5.209078229232603e-06,
+      "loss": 0.0068,
+      "step": 2528
+    },
+    {
+      "epoch": 0.5023338961167941,
+      "grad_norm": 0.7098839446706783,
+      "learning_rate": 5.205863474121972e-06,
+      "loss": 0.0152,
+      "step": 2529
+    },
+    {
+      "epoch": 0.5025325255735426,
+      "grad_norm": 0.3375041096332709,
+      "learning_rate": 5.2026486337634804e-06,
+      "loss": 0.0089,
+      "step": 2530
+    },
+    {
+      "epoch": 0.502731155030291,
+      "grad_norm": 0.41527798701142316,
+      "learning_rate": 5.199433709488394e-06,
+      "loss": 0.0141,
+      "step": 2531
+    },
+    {
+      "epoch": 0.5029297844870394,
+      "grad_norm": 0.39038459676259,
+      "learning_rate": 5.196218702628008e-06,
+      "loss": 0.0115,
+      "step": 2532
+    },
+    {
+      "epoch": 0.5031284139437878,
+      "grad_norm": 0.6528062096390664,
+      "learning_rate": 5.193003614513653e-06,
+      "loss": 0.0231,
+      "step": 2533
+    },
+    {
+      "epoch": 0.5033270434005362,
+      "grad_norm": 1.7917958287890257,
+      "learning_rate": 5.189788446476695e-06,
+      "loss": 0.0243,
+      "step": 2534
+    },
+    {
+      "epoch": 0.5035256728572848,
+      "grad_norm": 0.44912076622692515,
+      "learning_rate": 5.186573199848532e-06,
+      "loss": 0.0106,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5037243023140332,
+      "grad_norm": 0.40776135095164723,
+      "learning_rate": 5.183357875960592e-06,
+      "loss": 0.0122,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5039229317707816,
+      "grad_norm": 0.5810563884951965,
+      "learning_rate": 5.1801424761443404e-06,
+      "loss": 0.0164,
+      "step": 2537
+    },
+    {
+      "epoch": 0.50412156122753,
+      "grad_norm": 0.7568118365890487,
+      "learning_rate": 5.17692700173127e-06,
+      "loss": 0.0133,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5043201906842785,
+      "grad_norm": 0.42687995689269276,
+      "learning_rate": 5.173711454052905e-06,
+      "loss": 0.0095,
+      "step": 2539
+    },
+    {
+      "epoch": 0.5045188201410269,
+      "grad_norm": 0.4374548051838743,
+      "learning_rate": 5.170495834440802e-06,
+      "loss": 0.0121,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5047174495977753,
+      "grad_norm": 0.4622644074029037,
+      "learning_rate": 5.167280144226543e-06,
+      "loss": 0.0118,
+      "step": 2541
+    },
+    {
+      "epoch": 0.5049160790545237,
+      "grad_norm": 0.7577713982849161,
+      "learning_rate": 5.164064384741745e-06,
+      "loss": 0.0126,
+      "step": 2542
+    },
+    {
+      "epoch": 0.5051147085112723,
+      "grad_norm": 0.3726201408129837,
+      "learning_rate": 5.160848557318049e-06,
+      "loss": 0.0127,
+      "step": 2543
+    },
+    {
+      "epoch": 0.5053133379680207,
+      "grad_norm": 0.797286065504275,
+      "learning_rate": 5.157632663287126e-06,
+      "loss": 0.0126,
+      "step": 2544
+    },
+    {
+      "epoch": 0.5055119674247691,
+      "grad_norm": 0.4918743375069235,
+      "learning_rate": 5.1544167039806755e-06,
+      "loss": 0.0082,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5057105968815175,
+      "grad_norm": 1.8250694853729397,
+      "learning_rate": 5.1512006807304235e-06,
+      "loss": 0.0155,
+      "step": 2546
+    },
+    {
+      "epoch": 0.505909226338266,
+      "grad_norm": 0.6131857955146129,
+      "learning_rate": 5.1479845948681185e-06,
+      "loss": 0.0169,
+      "step": 2547
+    },
+    {
+      "epoch": 0.5061078557950144,
+      "grad_norm": 0.48035054079034195,
+      "learning_rate": 5.144768447725544e-06,
+      "loss": 0.009,
+      "step": 2548
+    },
+    {
+      "epoch": 0.5063064852517628,
+      "grad_norm": 0.3214285229524216,
+      "learning_rate": 5.1415522406344976e-06,
+      "loss": 0.0101,
+      "step": 2549
+    },
+    {
+      "epoch": 0.5065051147085112,
+      "grad_norm": 0.7075802214664444,
+      "learning_rate": 5.1383359749268114e-06,
+      "loss": 0.0066,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5067037441652598,
+      "grad_norm": 0.5521771461028792,
+      "learning_rate": 5.135119651934337e-06,
+      "loss": 0.0089,
+      "step": 2551
+    },
+    {
+      "epoch": 0.5069023736220082,
+      "grad_norm": 0.25160655768692114,
+      "learning_rate": 5.131903272988951e-06,
+      "loss": 0.0114,
+      "step": 2552
+    },
+    {
+      "epoch": 0.5071010030787566,
+      "grad_norm": 0.49214589306265416,
+      "learning_rate": 5.128686839422548e-06,
+      "loss": 0.0139,
+      "step": 2553
+    },
+    {
+      "epoch": 0.507299632535505,
+      "grad_norm": 0.3311557204020224,
+      "learning_rate": 5.125470352567057e-06,
+      "loss": 0.0085,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5074982619922535,
+      "grad_norm": 0.5102290741270068,
+      "learning_rate": 5.122253813754418e-06,
+      "loss": 0.0135,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5076968914490019,
+      "grad_norm": 0.47249684319774726,
+      "learning_rate": 5.119037224316596e-06,
+      "loss": 0.0142,
+      "step": 2556
+    },
+    {
+      "epoch": 0.5078955209057503,
+      "grad_norm": 0.605698552614918,
+      "learning_rate": 5.115820585585579e-06,
+      "loss": 0.0113,
+      "step": 2557
+    },
+    {
+      "epoch": 0.5080941503624987,
+      "grad_norm": 0.6910282577333509,
+      "learning_rate": 5.1126038988933745e-06,
+      "loss": 0.0098,
+      "step": 2558
+    },
+    {
+      "epoch": 0.5082927798192471,
+      "grad_norm": 0.3860296102305923,
+      "learning_rate": 5.109387165572007e-06,
+      "loss": 0.0149,
+      "step": 2559
+    },
+    {
+      "epoch": 0.5084914092759957,
+      "grad_norm": 0.47180358188811444,
+      "learning_rate": 5.106170386953524e-06,
+      "loss": 0.0099,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5086900387327441,
+      "grad_norm": 0.30576774774595783,
+      "learning_rate": 5.10295356436999e-06,
+      "loss": 0.0077,
+      "step": 2561
+    },
+    {
+      "epoch": 0.5088886681894925,
+      "grad_norm": 0.30760305826870893,
+      "learning_rate": 5.099736699153489e-06,
+      "loss": 0.0076,
+      "step": 2562
+    },
+    {
+      "epoch": 0.5090872976462409,
+      "grad_norm": 0.3838863165764435,
+      "learning_rate": 5.09651979263612e-06,
+      "loss": 0.0117,
+      "step": 2563
+    },
+    {
+      "epoch": 0.5092859271029894,
+      "grad_norm": 0.7031651252049763,
+      "learning_rate": 5.093302846150003e-06,
+      "loss": 0.0124,
+      "step": 2564
+    },
+    {
+      "epoch": 0.5094845565597378,
+      "grad_norm": 0.2976625692750354,
+      "learning_rate": 5.0900858610272686e-06,
+      "loss": 0.0081,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5096831860164862,
+      "grad_norm": 0.7678894575982894,
+      "learning_rate": 5.0868688386000705e-06,
+      "loss": 0.0082,
+      "step": 2566
+    },
+    {
+      "epoch": 0.5098818154732346,
+      "grad_norm": 1.202459097306564,
+      "learning_rate": 5.083651780200573e-06,
+      "loss": 0.0143,
+      "step": 2567
+    },
+    {
+      "epoch": 0.5100804449299832,
+      "grad_norm": 0.7918825373152566,
+      "learning_rate": 5.080434687160958e-06,
+      "loss": 0.0194,
+      "step": 2568
+    },
+    {
+      "epoch": 0.5102790743867316,
+      "grad_norm": 0.35999767545572886,
+      "learning_rate": 5.077217560813419e-06,
+      "loss": 0.0131,
+      "step": 2569
+    },
+    {
+      "epoch": 0.51047770384348,
+      "grad_norm": 0.6835356858570736,
+      "learning_rate": 5.074000402490166e-06,
+      "loss": 0.0125,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5106763333002284,
+      "grad_norm": 0.8376402761684941,
+      "learning_rate": 5.0707832135234196e-06,
+      "loss": 0.0111,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5108749627569769,
+      "grad_norm": 0.4974372233299198,
+      "learning_rate": 5.067565995245413e-06,
+      "loss": 0.0094,
+      "step": 2572
+    },
+    {
+      "epoch": 0.5110735922137253,
+      "grad_norm": 0.661553749690577,
+      "learning_rate": 5.0643487489883995e-06,
+      "loss": 0.0107,
+      "step": 2573
+    },
+    {
+      "epoch": 0.5112722216704737,
+      "grad_norm": 0.6776284432083003,
+      "learning_rate": 5.061131476084632e-06,
+      "loss": 0.0138,
+      "step": 2574
+    },
+    {
+      "epoch": 0.5114708511272221,
+      "grad_norm": 0.6037483195136601,
+      "learning_rate": 5.057914177866381e-06,
+      "loss": 0.0122,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5116694805839705,
+      "grad_norm": 0.5131467239719532,
+      "learning_rate": 5.054696855665928e-06,
+      "loss": 0.0103,
+      "step": 2576
+    },
+    {
+      "epoch": 0.5118681100407191,
+      "grad_norm": 0.43566722341066894,
+      "learning_rate": 5.051479510815561e-06,
+      "loss": 0.0098,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5120667394974675,
+      "grad_norm": 0.45003515001134503,
+      "learning_rate": 5.04826214464758e-06,
+      "loss": 0.0122,
+      "step": 2578
+    },
+    {
+      "epoch": 0.5122653689542159,
+      "grad_norm": 0.750722907248036,
+      "learning_rate": 5.0450447584942945e-06,
+      "loss": 0.0112,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5124639984109643,
+      "grad_norm": 0.6004993496158042,
+      "learning_rate": 5.041827353688018e-06,
+      "loss": 0.0185,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5126626278677128,
+      "grad_norm": 3.1872299490121687,
+      "learning_rate": 5.038609931561077e-06,
+      "loss": 0.0182,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5128612573244612,
+      "grad_norm": 0.57628867563776,
+      "learning_rate": 5.035392493445802e-06,
+      "loss": 0.0133,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5130598867812096,
+      "grad_norm": 0.4439739183814535,
+      "learning_rate": 5.03217504067453e-06,
+      "loss": 0.0083,
+      "step": 2583
+    },
+    {
+      "epoch": 0.513258516237958,
+      "grad_norm": 0.29526566381496966,
+      "learning_rate": 5.028957574579607e-06,
+      "loss": 0.0122,
+      "step": 2584
+    },
+    {
+      "epoch": 0.5134571456947066,
+      "grad_norm": 0.5866932164892713,
+      "learning_rate": 5.02574009649338e-06,
+      "loss": 0.0173,
+      "step": 2585
+    },
+    {
+      "epoch": 0.513655775151455,
+      "grad_norm": 0.6315522926249754,
+      "learning_rate": 5.0225226077482055e-06,
+      "loss": 0.0196,
+      "step": 2586
+    },
+    {
+      "epoch": 0.5138544046082034,
+      "grad_norm": 0.430607086857732,
+      "learning_rate": 5.019305109676443e-06,
+      "loss": 0.0118,
+      "step": 2587
+    },
+    {
+      "epoch": 0.5140530340649518,
+      "grad_norm": 0.6939527777771256,
+      "learning_rate": 5.016087603610454e-06,
+      "loss": 0.013,
+      "step": 2588
+    },
+    {
+      "epoch": 0.5142516635217003,
+      "grad_norm": 0.43910241298197344,
+      "learning_rate": 5.012870090882604e-06,
+      "loss": 0.0118,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5144502929784487,
+      "grad_norm": 0.49333214750546195,
+      "learning_rate": 5.00965257282526e-06,
+      "loss": 0.0104,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5146489224351971,
+      "grad_norm": 0.800628768602885,
+      "learning_rate": 5.006435050770797e-06,
+      "loss": 0.0112,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5148475518919455,
+      "grad_norm": 0.5851511459639136,
+      "learning_rate": 5.003217526051586e-06,
+      "loss": 0.012,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5150461813486941,
+      "grad_norm": 0.47041917943176503,
+      "learning_rate": 5e-06,
+      "loss": 0.0168,
+      "step": 2593
+    },
+    {
+      "epoch": 0.5152448108054425,
+      "grad_norm": 0.47959696982675354,
+      "learning_rate": 4.996782473948416e-06,
+      "loss": 0.0132,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5154434402621909,
+      "grad_norm": 0.47813958473253837,
+      "learning_rate": 4.993564949229204e-06,
+      "loss": 0.0108,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5156420697189393,
+      "grad_norm": 0.3852993385433787,
+      "learning_rate": 4.990347427174742e-06,
+      "loss": 0.0089,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5158406991756878,
+      "grad_norm": 0.6016579451534518,
+      "learning_rate": 4.987129909117398e-06,
+      "loss": 0.0175,
+      "step": 2597
+    },
+    {
+      "epoch": 0.5160393286324362,
+      "grad_norm": 0.45074106282002463,
+      "learning_rate": 4.983912396389547e-06,
+      "loss": 0.0094,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5162379580891846,
+      "grad_norm": 0.7101086980066497,
+      "learning_rate": 4.980694890323558e-06,
+      "loss": 0.0171,
+      "step": 2599
+    },
+    {
+      "epoch": 0.516436587545933,
+      "grad_norm": 0.4153502818151871,
+      "learning_rate": 4.977477392251794e-06,
+      "loss": 0.0176,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5166352170026814,
+      "grad_norm": 0.9128192784068005,
+      "learning_rate": 4.97425990350662e-06,
+      "loss": 0.0119,
+      "step": 2601
+    },
+    {
+      "epoch": 0.51683384645943,
+      "grad_norm": 0.532406897997197,
+      "learning_rate": 4.971042425420394e-06,
+      "loss": 0.0098,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5170324759161784,
+      "grad_norm": 0.6432723125325949,
+      "learning_rate": 4.96782495932547e-06,
+      "loss": 0.0125,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5172311053729268,
+      "grad_norm": 0.4441861132175126,
+      "learning_rate": 4.964607506554199e-06,
+      "loss": 0.0088,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5174297348296752,
+      "grad_norm": 0.6071744574706942,
+      "learning_rate": 4.961390068438926e-06,
+      "loss": 0.0125,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5176283642864237,
+      "grad_norm": 0.5144396277013178,
+      "learning_rate": 4.958172646311985e-06,
+      "loss": 0.0108,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5178269937431721,
+      "grad_norm": 0.27271582525812527,
+      "learning_rate": 4.954955241505709e-06,
+      "loss": 0.0057,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5180256231999205,
+      "grad_norm": 0.41602838022157057,
+      "learning_rate": 4.951737855352422e-06,
+      "loss": 0.014,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5182242526566689,
+      "grad_norm": 0.8548531908521733,
+      "learning_rate": 4.948520489184441e-06,
+      "loss": 0.0192,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5184228821134175,
+      "grad_norm": 0.7707419054286566,
+      "learning_rate": 4.945303144334074e-06,
+      "loss": 0.0156,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5186215115701659,
+      "grad_norm": 0.9002762306327531,
+      "learning_rate": 4.94208582213362e-06,
+      "loss": 0.0109,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5188201410269143,
+      "grad_norm": 0.5969255787289416,
+      "learning_rate": 4.9388685239153696e-06,
+      "loss": 0.013,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5190187704836627,
+      "grad_norm": 1.321061946411639,
+      "learning_rate": 4.935651251011602e-06,
+      "loss": 0.0265,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5192173999404112,
+      "grad_norm": 0.9544926651703668,
+      "learning_rate": 4.9324340047545874e-06,
+      "loss": 0.0232,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5194160293971596,
+      "grad_norm": 0.8724356174787311,
+      "learning_rate": 4.929216786476582e-06,
+      "loss": 0.0172,
+      "step": 2615
+    },
+    {
+      "epoch": 0.519614658853908,
+      "grad_norm": 0.5481496452758235,
+      "learning_rate": 4.925999597509836e-06,
+      "loss": 0.0105,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5198132883106564,
+      "grad_norm": 0.5924502586289119,
+      "learning_rate": 4.922782439186583e-06,
+      "loss": 0.0126,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5200119177674049,
+      "grad_norm": 1.097058049346646,
+      "learning_rate": 4.9195653128390436e-06,
+      "loss": 0.0194,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5202105472241534,
+      "grad_norm": 0.34572029206305155,
+      "learning_rate": 4.9163482197994275e-06,
+      "loss": 0.0114,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5204091766809018,
+      "grad_norm": 0.6478854101066456,
+      "learning_rate": 4.913131161399931e-06,
+      "loss": 0.0124,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5206078061376502,
+      "grad_norm": 0.8509089389165168,
+      "learning_rate": 4.909914138972733e-06,
+      "loss": 0.0152,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5208064355943987,
+      "grad_norm": 0.43045550772154256,
+      "learning_rate": 4.9066971538499985e-06,
+      "loss": 0.0058,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5210050650511471,
+      "grad_norm": 0.2811848823106481,
+      "learning_rate": 4.903480207363881e-06,
+      "loss": 0.0078,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5212036945078955,
+      "grad_norm": 0.17886282113428878,
+      "learning_rate": 4.900263300846512e-06,
+      "loss": 0.0045,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5214023239646439,
+      "grad_norm": 1.0905060510323827,
+      "learning_rate": 4.89704643563001e-06,
+      "loss": 0.021,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5216009534213923,
+      "grad_norm": 0.52942449810658,
+      "learning_rate": 4.893829613046476e-06,
+      "loss": 0.0105,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5217995828781409,
+      "grad_norm": 0.38872032429287773,
+      "learning_rate": 4.890612834427994e-06,
+      "loss": 0.0119,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5219982123348893,
+      "grad_norm": 0.4144058112551697,
+      "learning_rate": 4.887396101106627e-06,
+      "loss": 0.0122,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5221968417916377,
+      "grad_norm": 0.7047127790035546,
+      "learning_rate": 4.884179414414423e-06,
+      "loss": 0.013,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5223954712483861,
+      "grad_norm": 0.41046615550881144,
+      "learning_rate": 4.880962775683406e-06,
+      "loss": 0.0143,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5225941007051346,
+      "grad_norm": 0.2756154625309793,
+      "learning_rate": 4.877746186245586e-06,
+      "loss": 0.0114,
+      "step": 2631
+    },
+    {
+      "epoch": 0.522792730161883,
+      "grad_norm": 0.8783246622943536,
+      "learning_rate": 4.874529647432946e-06,
+      "loss": 0.0111,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5229913596186314,
+      "grad_norm": 0.35351879117875357,
+      "learning_rate": 4.871313160577453e-06,
+      "loss": 0.0114,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5231899890753798,
+      "grad_norm": 0.3497304664011684,
+      "learning_rate": 4.868096727011052e-06,
+      "loss": 0.0117,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5233886185321284,
+      "grad_norm": 0.38460261605226964,
+      "learning_rate": 4.864880348065665e-06,
+      "loss": 0.0099,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5235872479888768,
+      "grad_norm": 0.34385257857225565,
+      "learning_rate": 4.86166402507319e-06,
+      "loss": 0.0131,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5237858774456252,
+      "grad_norm": 0.7154889328263488,
+      "learning_rate": 4.858447759365504e-06,
+      "loss": 0.0185,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5239845069023736,
+      "grad_norm": 0.45589891363247864,
+      "learning_rate": 4.855231552274459e-06,
+      "loss": 0.0089,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5241831363591221,
+      "grad_norm": 0.3846830282655784,
+      "learning_rate": 4.852015405131882e-06,
+      "loss": 0.018,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5243817658158705,
+      "grad_norm": 0.2940469069583859,
+      "learning_rate": 4.848799319269578e-06,
+      "loss": 0.0064,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5245803952726189,
+      "grad_norm": 0.2149661207674454,
+      "learning_rate": 4.845583296019325e-06,
+      "loss": 0.0061,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5247790247293673,
+      "grad_norm": 0.3446058201626362,
+      "learning_rate": 4.8423673367128745e-06,
+      "loss": 0.0072,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5249776541861157,
+      "grad_norm": 0.9384941741122217,
+      "learning_rate": 4.839151442681952e-06,
+      "loss": 0.014,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5251762836428643,
+      "grad_norm": 0.7648356533107046,
+      "learning_rate": 4.835935615258257e-06,
+      "loss": 0.0112,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5253749130996127,
+      "grad_norm": 0.5947850726160407,
+      "learning_rate": 4.832719855773458e-06,
+      "loss": 0.0237,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5255735425563611,
+      "grad_norm": 0.5069947415968155,
+      "learning_rate": 4.8295041655592e-06,
+      "loss": 0.019,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5257721720131096,
+      "grad_norm": 0.530771872211901,
+      "learning_rate": 4.826288545947095e-06,
+      "loss": 0.0167,
+      "step": 2647
+    },
+    {
+      "epoch": 0.525970801469858,
+      "grad_norm": 0.7343719370370166,
+      "learning_rate": 4.82307299826873e-06,
+      "loss": 0.0153,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5261694309266064,
+      "grad_norm": 0.43053313989705466,
+      "learning_rate": 4.81985752385566e-06,
+      "loss": 0.0113,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5263680603833548,
+      "grad_norm": 1.0486479094087207,
+      "learning_rate": 4.816642124039408e-06,
+      "loss": 0.0225,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5265666898401032,
+      "grad_norm": 1.204001697819146,
+      "learning_rate": 4.813426800151469e-06,
+      "loss": 0.0119,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5267653192968518,
+      "grad_norm": 0.38383932212493976,
+      "learning_rate": 4.8102115535233054e-06,
+      "loss": 0.0108,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5269639487536002,
+      "grad_norm": 0.7521314041717428,
+      "learning_rate": 4.806996385486349e-06,
+      "loss": 0.0115,
+      "step": 2653
+    },
+    {
+      "epoch": 0.5271625782103486,
+      "grad_norm": 0.35629755731123935,
+      "learning_rate": 4.803781297371995e-06,
+      "loss": 0.0073,
+      "step": 2654
+    },
+    {
+      "epoch": 0.527361207667097,
+      "grad_norm": 0.4526855571654173,
+      "learning_rate": 4.8005662905116085e-06,
+      "loss": 0.0092,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5275598371238455,
+      "grad_norm": 0.3182080909457293,
+      "learning_rate": 4.797351366236522e-06,
+      "loss": 0.0132,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5277584665805939,
+      "grad_norm": 0.6049331198936719,
+      "learning_rate": 4.794136525878032e-06,
+      "loss": 0.016,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5279570960373423,
+      "grad_norm": 0.4754537156663623,
+      "learning_rate": 4.7909217707673984e-06,
+      "loss": 0.0168,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5281557254940907,
+      "grad_norm": 0.998499211825424,
+      "learning_rate": 4.787707102235852e-06,
+      "loss": 0.0176,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5283543549508392,
+      "grad_norm": 0.46739699878852936,
+      "learning_rate": 4.784492521614582e-06,
+      "loss": 0.01,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5285529844075877,
+      "grad_norm": 0.4320350647544851,
+      "learning_rate": 4.781278030234742e-06,
+      "loss": 0.0078,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5287516138643361,
+      "grad_norm": 0.6016783731374997,
+      "learning_rate": 4.778063629427451e-06,
+      "loss": 0.0127,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5289502433210845,
+      "grad_norm": 0.49313103783452733,
+      "learning_rate": 4.77484932052379e-06,
+      "loss": 0.0146,
+      "step": 2663
+    },
+    {
+      "epoch": 0.529148872777833,
+      "grad_norm": 0.40985189124421423,
+      "learning_rate": 4.771635104854799e-06,
+      "loss": 0.0138,
+      "step": 2664
+    },
+    {
+      "epoch": 0.5293475022345814,
+      "grad_norm": 0.3135781617489447,
+      "learning_rate": 4.768420983751481e-06,
+      "loss": 0.0105,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5295461316913298,
+      "grad_norm": 0.41689021986146035,
+      "learning_rate": 4.765206958544803e-06,
+      "loss": 0.0154,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5297447611480782,
+      "grad_norm": 0.2399269526721442,
+      "learning_rate": 4.761993030565688e-06,
+      "loss": 0.0076,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5299433906048266,
+      "grad_norm": 0.49800855497998864,
+      "learning_rate": 4.758779201145022e-06,
+      "loss": 0.02,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5301420200615752,
+      "grad_norm": 0.3540794930683414,
+      "learning_rate": 4.755565471613646e-06,
+      "loss": 0.0074,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5303406495183236,
+      "grad_norm": 0.37327188769326325,
+      "learning_rate": 4.752351843302364e-06,
+      "loss": 0.0093,
+      "step": 2670
+    },
+    {
+      "epoch": 0.530539278975072,
+      "grad_norm": 0.6218524242302055,
+      "learning_rate": 4.749138317541936e-06,
+      "loss": 0.0114,
+      "step": 2671
+    },
+    {
+      "epoch": 0.5307379084318204,
+      "grad_norm": 0.47629545880771196,
+      "learning_rate": 4.745924895663078e-06,
+      "loss": 0.0095,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5309365378885689,
+      "grad_norm": 0.41948688036972176,
+      "learning_rate": 4.742711578996467e-06,
+      "loss": 0.0128,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5311351673453173,
+      "grad_norm": 0.26707306675736925,
+      "learning_rate": 4.739498368872734e-06,
+      "loss": 0.0114,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5313337968020657,
+      "grad_norm": 0.4462577977563272,
+      "learning_rate": 4.7362852666224654e-06,
+      "loss": 0.022,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5315324262588141,
+      "grad_norm": 0.42752531113748565,
+      "learning_rate": 4.733072273576204e-06,
+      "loss": 0.0128,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5317310557155627,
+      "grad_norm": 0.1988579553958363,
+      "learning_rate": 4.729859391064447e-06,
+      "loss": 0.0066,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5319296851723111,
+      "grad_norm": 0.3623897142655911,
+      "learning_rate": 4.726646620417646e-06,
+      "loss": 0.0082,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5321283146290595,
+      "grad_norm": 0.5471633235503327,
+      "learning_rate": 4.723433962966204e-06,
+      "loss": 0.0099,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5323269440858079,
+      "grad_norm": 0.30009321328868405,
+      "learning_rate": 4.720221420040483e-06,
+      "loss": 0.0084,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5325255735425564,
+      "grad_norm": 0.36399418358138586,
+      "learning_rate": 4.71700899297079e-06,
+      "loss": 0.0081,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5327242029993048,
+      "grad_norm": 0.4970147699892617,
+      "learning_rate": 4.7137966830873905e-06,
+      "loss": 0.0113,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5329228324560532,
+      "grad_norm": 0.5953544717285123,
+      "learning_rate": 4.710584491720496e-06,
+      "loss": 0.0182,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5331214619128016,
+      "grad_norm": 0.4802266373130576,
+      "learning_rate": 4.707372420200275e-06,
+      "loss": 0.0124,
+      "step": 2684
+    },
+    {
+      "epoch": 0.53332009136955,
+      "grad_norm": 0.6689446433586343,
+      "learning_rate": 4.7041604698568436e-06,
+      "loss": 0.0196,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5335187208262986,
+      "grad_norm": 0.5247125265711512,
+      "learning_rate": 4.700948642020263e-06,
+      "loss": 0.0108,
+      "step": 2686
+    },
+    {
+      "epoch": 0.533717350283047,
+      "grad_norm": 0.4728274310812671,
+      "learning_rate": 4.6977369380205514e-06,
+      "loss": 0.0123,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5339159797397954,
+      "grad_norm": 0.60501199620917,
+      "learning_rate": 4.694525359187671e-06,
+      "loss": 0.0071,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5341146091965439,
+      "grad_norm": 0.36671802535764086,
+      "learning_rate": 4.691313906851534e-06,
+      "loss": 0.0119,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5343132386532923,
+      "grad_norm": 1.0242303491155098,
+      "learning_rate": 4.688102582341997e-06,
+      "loss": 0.0172,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5345118681100407,
+      "grad_norm": 0.7058136287291272,
+      "learning_rate": 4.684891386988869e-06,
+      "loss": 0.014,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5347104975667891,
+      "grad_norm": 1.1395881555357235,
+      "learning_rate": 4.681680322121903e-06,
+      "loss": 0.0197,
+      "step": 2692
+    },
+    {
+      "epoch": 0.5349091270235375,
+      "grad_norm": 0.8146146171242227,
+      "learning_rate": 4.678469389070797e-06,
+      "loss": 0.0135,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5351077564802861,
+      "grad_norm": 0.4681663478402444,
+      "learning_rate": 4.675258589165194e-06,
+      "loss": 0.0061,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5353063859370345,
+      "grad_norm": 0.7888374699099346,
+      "learning_rate": 4.672047923734685e-06,
+      "loss": 0.0126,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5355050153937829,
+      "grad_norm": 0.26698990733974676,
+      "learning_rate": 4.668837394108801e-06,
+      "loss": 0.0081,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5357036448505313,
+      "grad_norm": 0.6481355354112126,
+      "learning_rate": 4.665627001617021e-06,
+      "loss": 0.0106,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5359022743072798,
+      "grad_norm": 0.9141985443318152,
+      "learning_rate": 4.662416747588765e-06,
+      "loss": 0.019,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5361009037640282,
+      "grad_norm": 0.345436456845145,
+      "learning_rate": 4.6592066333533966e-06,
+      "loss": 0.0101,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5362995332207766,
+      "grad_norm": 0.5340280696377245,
+      "learning_rate": 4.6559966602402195e-06,
+      "loss": 0.0132,
+      "step": 2700
+    },
+    {
+      "epoch": 0.536498162677525,
+      "grad_norm": 0.495489992727139,
+      "learning_rate": 4.652786829578482e-06,
+      "loss": 0.0108,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5366967921342735,
+      "grad_norm": 0.48163074793680033,
+      "learning_rate": 4.649577142697369e-06,
+      "loss": 0.0089,
+      "step": 2702
+    },
+    {
+      "epoch": 0.536895421591022,
+      "grad_norm": 0.6202192336463302,
+      "learning_rate": 4.6463676009260115e-06,
+      "loss": 0.0128,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5370940510477704,
+      "grad_norm": 0.8438747669064174,
+      "learning_rate": 4.643158205593475e-06,
+      "loss": 0.0207,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5372926805045188,
+      "grad_norm": 0.5444604148537907,
+      "learning_rate": 4.639948958028769e-06,
+      "loss": 0.011,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5374913099612673,
+      "grad_norm": 0.40458432983482345,
+      "learning_rate": 4.636739859560839e-06,
+      "loss": 0.0137,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5376899394180157,
+      "grad_norm": 0.38258143370225456,
+      "learning_rate": 4.633530911518569e-06,
+      "loss": 0.0121,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5378885688747641,
+      "grad_norm": 0.3338515448391374,
+      "learning_rate": 4.63032211523078e-06,
+      "loss": 0.0081,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5380871983315125,
+      "grad_norm": 0.2721756551155944,
+      "learning_rate": 4.627113472026235e-06,
+      "loss": 0.0114,
+      "step": 2709
+    },
+    {
+      "epoch": 0.538285827788261,
+      "grad_norm": 0.8199742148753297,
+      "learning_rate": 4.623904983233628e-06,
+      "loss": 0.0194,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5384844572450095,
+      "grad_norm": 0.6501960596074884,
+      "learning_rate": 4.6206966501815895e-06,
+      "loss": 0.0105,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5386830867017579,
+      "grad_norm": 0.3113077005159778,
+      "learning_rate": 4.617488474198689e-06,
+      "loss": 0.0066,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5388817161585063,
+      "grad_norm": 0.46285719527025915,
+      "learning_rate": 4.614280456613428e-06,
+      "loss": 0.015,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5390803456152548,
+      "grad_norm": 0.9043967381501636,
+      "learning_rate": 4.6110725987542436e-06,
+      "loss": 0.0161,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5392789750720032,
+      "grad_norm": 0.6074770026179739,
+      "learning_rate": 4.607864901949506e-06,
+      "loss": 0.0085,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5394776045287516,
+      "grad_norm": 0.5580886715434685,
+      "learning_rate": 4.6046573675275204e-06,
+      "loss": 0.0124,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5396762339855,
+      "grad_norm": 0.4824079804144595,
+      "learning_rate": 4.601449996816524e-06,
+      "loss": 0.0099,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5398748634422484,
+      "grad_norm": 0.48189458946683134,
+      "learning_rate": 4.598242791144684e-06,
+      "loss": 0.0126,
+      "step": 2718
+    },
+    {
+      "epoch": 0.540073492898997,
+      "grad_norm": 1.2183320489020404,
+      "learning_rate": 4.5950357518401015e-06,
+      "loss": 0.0142,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5402721223557454,
+      "grad_norm": 0.30529970276531804,
+      "learning_rate": 4.591828880230809e-06,
+      "loss": 0.0101,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5404707518124938,
+      "grad_norm": 0.23725652114441975,
+      "learning_rate": 4.588622177644769e-06,
+      "loss": 0.0052,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5406693812692422,
+      "grad_norm": 0.49526465559783595,
+      "learning_rate": 4.585415645409872e-06,
+      "loss": 0.0101,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5408680107259907,
+      "grad_norm": 0.629120686573954,
+      "learning_rate": 4.5822092848539415e-06,
+      "loss": 0.0178,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5410666401827391,
+      "grad_norm": 0.6034085054382206,
+      "learning_rate": 4.579003097304728e-06,
+      "loss": 0.0152,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5412652696394875,
+      "grad_norm": 0.32942180161724255,
+      "learning_rate": 4.575797084089912e-06,
+      "loss": 0.0093,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5414638990962359,
+      "grad_norm": 0.42192709837470616,
+      "learning_rate": 4.572591246537097e-06,
+      "loss": 0.0112,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5416625285529844,
+      "grad_norm": 0.6217230182883882,
+      "learning_rate": 4.569385585973818e-06,
+      "loss": 0.01,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5418611580097329,
+      "grad_norm": 0.6470622146574528,
+      "learning_rate": 4.566180103727538e-06,
+      "loss": 0.0172,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5420597874664813,
+      "grad_norm": 0.15106597972191477,
+      "learning_rate": 4.562974801125642e-06,
+      "loss": 0.0053,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5422584169232297,
+      "grad_norm": 0.6869662187626556,
+      "learning_rate": 4.559769679495443e-06,
+      "loss": 0.0138,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5424570463799782,
+      "grad_norm": 0.34343413551532936,
+      "learning_rate": 4.55656474016418e-06,
+      "loss": 0.0077,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5426556758367266,
+      "grad_norm": 0.6162031016897686,
+      "learning_rate": 4.553359984459012e-06,
+      "loss": 0.0146,
+      "step": 2732
+    },
+    {
+      "epoch": 0.542854305293475,
+      "grad_norm": 0.33775023474661575,
+      "learning_rate": 4.550155413707028e-06,
+      "loss": 0.0112,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5430529347502234,
+      "grad_norm": 0.5286312992072117,
+      "learning_rate": 4.546951029235237e-06,
+      "loss": 0.0089,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5432515642069718,
+      "grad_norm": 0.44205639734611607,
+      "learning_rate": 4.543746832370572e-06,
+      "loss": 0.0101,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5434501936637204,
+      "grad_norm": 0.7046118783607311,
+      "learning_rate": 4.540542824439885e-06,
+      "loss": 0.0133,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5436488231204688,
+      "grad_norm": 0.2920099315721433,
+      "learning_rate": 4.5373390067699555e-06,
+      "loss": 0.0075,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5438474525772172,
+      "grad_norm": 0.40661921340738155,
+      "learning_rate": 4.53413538068748e-06,
+      "loss": 0.0118,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5440460820339657,
+      "grad_norm": 0.7267749331431869,
+      "learning_rate": 4.530931947519076e-06,
+      "loss": 0.0148,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5442447114907141,
+      "grad_norm": 0.3581174389427978,
+      "learning_rate": 4.527728708591283e-06,
+      "loss": 0.0107,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5444433409474625,
+      "grad_norm": 0.41211789962071976,
+      "learning_rate": 4.524525665230559e-06,
+      "loss": 0.0129,
+      "step": 2741
+    },
+    {
+      "epoch": 0.5446419704042109,
+      "grad_norm": 0.9758329205888799,
+      "learning_rate": 4.521322818763281e-06,
+      "loss": 0.0179,
+      "step": 2742
+    },
+    {
+      "epoch": 0.5448405998609593,
+      "grad_norm": 0.8121007154836791,
+      "learning_rate": 4.518120170515744e-06,
+      "loss": 0.018,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5450392293177078,
+      "grad_norm": 0.2556833343570392,
+      "learning_rate": 4.514917721814163e-06,
+      "loss": 0.0062,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5452378587744563,
+      "grad_norm": 0.49149727060895204,
+      "learning_rate": 4.5117154739846665e-06,
+      "loss": 0.0176,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5454364882312047,
+      "grad_norm": 1.0477647176321618,
+      "learning_rate": 4.5085134283533035e-06,
+      "loss": 0.0168,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5456351176879531,
+      "grad_norm": 0.3199467470120506,
+      "learning_rate": 4.505311586246037e-06,
+      "loss": 0.0087,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5458337471447016,
+      "grad_norm": 0.4189748938840971,
+      "learning_rate": 4.502109948988748e-06,
+      "loss": 0.0122,
+      "step": 2748
+    },
+    {
+      "epoch": 0.54603237660145,
+      "grad_norm": 0.43642675146362775,
+      "learning_rate": 4.498908517907232e-06,
+      "loss": 0.0065,
+      "step": 2749
+    },
+    {
+      "epoch": 0.5462310060581984,
+      "grad_norm": 1.5799130468680098,
+      "learning_rate": 4.4957072943271965e-06,
+      "loss": 0.0129,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5464296355149468,
+      "grad_norm": 0.3761584834352775,
+      "learning_rate": 4.492506279574262e-06,
+      "loss": 0.0124,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5466282649716953,
+      "grad_norm": 0.7976727590944505,
+      "learning_rate": 4.4893054749739715e-06,
+      "loss": 0.0131,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5468268944284438,
+      "grad_norm": 1.038660703724904,
+      "learning_rate": 4.4861048818517725e-06,
+      "loss": 0.0164,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5470255238851922,
+      "grad_norm": 0.8037713236955946,
+      "learning_rate": 4.482904501533027e-06,
+      "loss": 0.0148,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5472241533419406,
+      "grad_norm": 0.3649529151254279,
+      "learning_rate": 4.47970433534301e-06,
+      "loss": 0.0109,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5474227827986891,
+      "grad_norm": 0.539372194994621,
+      "learning_rate": 4.476504384606906e-06,
+      "loss": 0.0128,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5476214122554375,
+      "grad_norm": 0.7258264788596848,
+      "learning_rate": 4.473304650649812e-06,
+      "loss": 0.0146,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5478200417121859,
+      "grad_norm": 0.28150530699276105,
+      "learning_rate": 4.470105134796734e-06,
+      "loss": 0.0065,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5480186711689343,
+      "grad_norm": 0.7482265228868586,
+      "learning_rate": 4.466905838372591e-06,
+      "loss": 0.0179,
+      "step": 2759
+    },
+    {
+      "epoch": 0.5482173006256827,
+      "grad_norm": 0.6119275773400306,
+      "learning_rate": 4.463706762702205e-06,
+      "loss": 0.0176,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5484159300824313,
+      "grad_norm": 0.4938248337453184,
+      "learning_rate": 4.460507909110312e-06,
+      "loss": 0.0106,
+      "step": 2761
+    },
+    {
+      "epoch": 0.5486145595391797,
+      "grad_norm": 0.27287982790646315,
+      "learning_rate": 4.457309278921554e-06,
+      "loss": 0.0054,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5488131889959281,
+      "grad_norm": 0.7541141348539235,
+      "learning_rate": 4.4541108734604795e-06,
+      "loss": 0.0108,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5490118184526765,
+      "grad_norm": 0.7194720755343721,
+      "learning_rate": 4.450912694051546e-06,
+      "loss": 0.0131,
+      "step": 2764
+    },
+    {
+      "epoch": 0.549210447909425,
+      "grad_norm": 0.36732971751956595,
+      "learning_rate": 4.447714742019115e-06,
+      "loss": 0.0114,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5494090773661734,
+      "grad_norm": 0.6386170728830878,
+      "learning_rate": 4.444517018687457e-06,
+      "loss": 0.009,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5496077068229218,
+      "grad_norm": 0.39169752604182223,
+      "learning_rate": 4.441319525380745e-06,
+      "loss": 0.0133,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5498063362796702,
+      "grad_norm": 0.569101509077439,
+      "learning_rate": 4.438122263423059e-06,
+      "loss": 0.0131,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5500049657364187,
+      "grad_norm": 0.7374889233619399,
+      "learning_rate": 4.434925234138381e-06,
+      "loss": 0.0154,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5502035951931672,
+      "grad_norm": 0.27469994570416645,
+      "learning_rate": 4.431728438850597e-06,
+      "loss": 0.01,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5504022246499156,
+      "grad_norm": 0.5480817776491356,
+      "learning_rate": 4.4285318788834976e-06,
+      "loss": 0.0112,
+      "step": 2771
+    },
+    {
+      "epoch": 0.550600854106664,
+      "grad_norm": 0.4678729381775268,
+      "learning_rate": 4.425335555560773e-06,
+      "loss": 0.0112,
+      "step": 2772
+    },
+    {
+      "epoch": 0.5507994835634125,
+      "grad_norm": 0.6929526581318076,
+      "learning_rate": 4.422139470206024e-06,
+      "loss": 0.0147,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5509981130201609,
+      "grad_norm": 0.5590965545680023,
+      "learning_rate": 4.4189436241427395e-06,
+      "loss": 0.0128,
+      "step": 2774
+    },
+    {
+      "epoch": 0.5511967424769093,
+      "grad_norm": 0.4299925224605084,
+      "learning_rate": 4.415748018694317e-06,
+      "loss": 0.0078,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5513953719336577,
+      "grad_norm": 0.30426919330239416,
+      "learning_rate": 4.412552655184055e-06,
+      "loss": 0.0071,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5515940013904062,
+      "grad_norm": 1.7960268104440869,
+      "learning_rate": 4.409357534935151e-06,
+      "loss": 0.0276,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5517926308471547,
+      "grad_norm": 2.6449701275835187,
+      "learning_rate": 4.4061626592707e-06,
+      "loss": 0.0182,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5519912603039031,
+      "grad_norm": 0.22507808021610604,
+      "learning_rate": 4.4029680295136975e-06,
+      "loss": 0.0056,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5521898897606515,
+      "grad_norm": 0.7832386144210031,
+      "learning_rate": 4.399773646987036e-06,
+      "loss": 0.0183,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5523885192174,
+      "grad_norm": 0.632793978123542,
+      "learning_rate": 4.396579513013506e-06,
+      "loss": 0.0148,
+      "step": 2781
+    },
+    {
+      "epoch": 0.5525871486741484,
+      "grad_norm": 0.5274197472069998,
+      "learning_rate": 4.393385628915795e-06,
+      "loss": 0.0132,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5527857781308968,
+      "grad_norm": 0.6239929821382354,
+      "learning_rate": 4.390191996016488e-06,
+      "loss": 0.0139,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5529844075876452,
+      "grad_norm": 0.5616827331201757,
+      "learning_rate": 4.386998615638064e-06,
+      "loss": 0.0099,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5531830370443936,
+      "grad_norm": 0.5048749306077575,
+      "learning_rate": 4.383805489102901e-06,
+      "loss": 0.0098,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5533816665011421,
+      "grad_norm": 0.3937045403697834,
+      "learning_rate": 4.380612617733267e-06,
+      "loss": 0.0106,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5535802959578906,
+      "grad_norm": 0.4297000120260985,
+      "learning_rate": 4.3774200028513275e-06,
+      "loss": 0.0085,
+      "step": 2787
+    },
+    {
+      "epoch": 0.553778925414639,
+      "grad_norm": 0.7721765597312535,
+      "learning_rate": 4.374227645779142e-06,
+      "loss": 0.0146,
+      "step": 2788
+    },
+    {
+      "epoch": 0.5539775548713874,
+      "grad_norm": 0.6219951031088091,
+      "learning_rate": 4.371035547838661e-06,
+      "loss": 0.0123,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5541761843281359,
+      "grad_norm": 0.6784123591321396,
+      "learning_rate": 4.36784371035173e-06,
+      "loss": 0.0119,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5543748137848843,
+      "grad_norm": 0.4407590870450644,
+      "learning_rate": 4.364652134640085e-06,
+      "loss": 0.0105,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5545734432416327,
+      "grad_norm": 0.5200832710265008,
+      "learning_rate": 4.361460822025356e-06,
+      "loss": 0.01,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5547720726983811,
+      "grad_norm": 0.5895406136146861,
+      "learning_rate": 4.35826977382906e-06,
+      "loss": 0.0082,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5549707021551296,
+      "grad_norm": 0.3137853635557372,
+      "learning_rate": 4.355078991372609e-06,
+      "loss": 0.0062,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5551693316118781,
+      "grad_norm": 0.3276947209644395,
+      "learning_rate": 4.351888475977302e-06,
+      "loss": 0.0062,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5553679610686265,
+      "grad_norm": 0.49943742040860356,
+      "learning_rate": 4.348698228964327e-06,
+      "loss": 0.0077,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5555665905253749,
+      "grad_norm": 0.46359295428980324,
+      "learning_rate": 4.345508251654765e-06,
+      "loss": 0.0068,
+      "step": 2797
+    },
+    {
+      "epoch": 0.5557652199821234,
+      "grad_norm": 1.1215529581908548,
+      "learning_rate": 4.34231854536958e-06,
+      "loss": 0.0217,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5559638494388718,
+      "grad_norm": 0.6207216506521158,
+      "learning_rate": 4.339129111429625e-06,
+      "loss": 0.0096,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5561624788956202,
+      "grad_norm": 0.3028484585994268,
+      "learning_rate": 4.335939951155644e-06,
+      "loss": 0.0053,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5563611083523686,
+      "grad_norm": 1.030312295449838,
+      "learning_rate": 4.332751065868264e-06,
+      "loss": 0.0145,
+      "step": 2801
+    },
+    {
+      "epoch": 0.556559737809117,
+      "grad_norm": 0.4829944470180176,
+      "learning_rate": 4.329562456888e-06,
+      "loss": 0.0117,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5567583672658655,
+      "grad_norm": 0.3015442911784513,
+      "learning_rate": 4.32637412553525e-06,
+      "loss": 0.0066,
+      "step": 2803
+    },
+    {
+      "epoch": 0.556956996722614,
+      "grad_norm": 0.49410995263115004,
+      "learning_rate": 4.323186073130302e-06,
+      "loss": 0.0134,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5571556261793624,
+      "grad_norm": 1.1059775844917823,
+      "learning_rate": 4.319998300993322e-06,
+      "loss": 0.0167,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5573542556361109,
+      "grad_norm": 0.5759009436326822,
+      "learning_rate": 4.316810810444365e-06,
+      "loss": 0.0121,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5575528850928593,
+      "grad_norm": 0.41776379254655316,
+      "learning_rate": 4.313623602803367e-06,
+      "loss": 0.0048,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5577515145496077,
+      "grad_norm": 0.5034867323356977,
+      "learning_rate": 4.310436679390147e-06,
+      "loss": 0.0121,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5579501440063561,
+      "grad_norm": 0.41318019364973024,
+      "learning_rate": 4.307250041524408e-06,
+      "loss": 0.0064,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5581487734631045,
+      "grad_norm": 1.0091161435064295,
+      "learning_rate": 4.304063690525734e-06,
+      "loss": 0.0137,
+      "step": 2810
+    },
+    {
+      "epoch": 0.558347402919853,
+      "grad_norm": 0.6276179334811938,
+      "learning_rate": 4.300877627713588e-06,
+      "loss": 0.0113,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5585460323766015,
+      "grad_norm": 0.8074924839212643,
+      "learning_rate": 4.297691854407317e-06,
+      "loss": 0.0175,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5587446618333499,
+      "grad_norm": 1.2308453281661025,
+      "learning_rate": 4.294506371926145e-06,
+      "loss": 0.0256,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5589432912900983,
+      "grad_norm": 0.8890857016385656,
+      "learning_rate": 4.291321181589179e-06,
+      "loss": 0.0138,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5591419207468468,
+      "grad_norm": 0.42316267167957217,
+      "learning_rate": 4.288136284715399e-06,
+      "loss": 0.0061,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5593405502035952,
+      "grad_norm": 0.9297151487475857,
+      "learning_rate": 4.284951682623674e-06,
+      "loss": 0.0149,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5595391796603436,
+      "grad_norm": 0.4544147000009162,
+      "learning_rate": 4.281767376632739e-06,
+      "loss": 0.0071,
+      "step": 2817
+    },
+    {
+      "epoch": 0.559737809117092,
+      "grad_norm": 0.6237208093686428,
+      "learning_rate": 4.278583368061216e-06,
+      "loss": 0.0225,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5599364385738405,
+      "grad_norm": 0.39156008750074794,
+      "learning_rate": 4.275399658227596e-06,
+      "loss": 0.0123,
+      "step": 2819
+    },
+    {
+      "epoch": 0.560135068030589,
+      "grad_norm": 0.41234897144867805,
+      "learning_rate": 4.272216248450253e-06,
+      "loss": 0.0072,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5603336974873374,
+      "grad_norm": 0.4726168304536307,
+      "learning_rate": 4.2690331400474335e-06,
+      "loss": 0.0095,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5605323269440858,
+      "grad_norm": 0.36137404100034154,
+      "learning_rate": 4.265850334337258e-06,
+      "loss": 0.0106,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5607309564008343,
+      "grad_norm": 0.41668059273849295,
+      "learning_rate": 4.262667832637724e-06,
+      "loss": 0.0085,
+      "step": 2823
+    },
+    {
+      "epoch": 0.5609295858575827,
+      "grad_norm": 0.547862940561988,
+      "learning_rate": 4.2594856362667015e-06,
+      "loss": 0.0129,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5611282153143311,
+      "grad_norm": 0.40441643120996035,
+      "learning_rate": 4.256303746541936e-06,
+      "loss": 0.0102,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5613268447710795,
+      "grad_norm": 0.5954924239361523,
+      "learning_rate": 4.253122164781043e-06,
+      "loss": 0.0143,
+      "step": 2826
+    },
+    {
+      "epoch": 0.561525474227828,
+      "grad_norm": 0.7813919479034077,
+      "learning_rate": 4.249940892301514e-06,
+      "loss": 0.009,
+      "step": 2827
+    },
+    {
+      "epoch": 0.5617241036845764,
+      "grad_norm": 0.4565410720159291,
+      "learning_rate": 4.246759930420711e-06,
+      "loss": 0.0124,
+      "step": 2828
+    },
+    {
+      "epoch": 0.5619227331413249,
+      "grad_norm": 0.5694645470193745,
+      "learning_rate": 4.243579280455867e-06,
+      "loss": 0.0108,
+      "step": 2829
+    },
+    {
+      "epoch": 0.5621213625980733,
+      "grad_norm": 0.8304296127324543,
+      "learning_rate": 4.240398943724085e-06,
+      "loss": 0.0101,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5623199920548217,
+      "grad_norm": 0.7021520729713497,
+      "learning_rate": 4.237218921542339e-06,
+      "loss": 0.0177,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5625186215115702,
+      "grad_norm": 0.6305791789892291,
+      "learning_rate": 4.234039215227474e-06,
+      "loss": 0.009,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5627172509683186,
+      "grad_norm": 0.9230478228209408,
+      "learning_rate": 4.230859826096203e-06,
+      "loss": 0.0147,
+      "step": 2833
+    },
+    {
+      "epoch": 0.562915880425067,
+      "grad_norm": 0.4300837811713559,
+      "learning_rate": 4.2276807554651074e-06,
+      "loss": 0.0099,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5631145098818154,
+      "grad_norm": 0.5975455251314918,
+      "learning_rate": 4.224502004650639e-06,
+      "loss": 0.0138,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5633131393385639,
+      "grad_norm": 0.5208038837005452,
+      "learning_rate": 4.221323574969113e-06,
+      "loss": 0.0129,
+      "step": 2836
+    },
+    {
+      "epoch": 0.5635117687953124,
+      "grad_norm": 0.4113017440579882,
+      "learning_rate": 4.218145467736715e-06,
+      "loss": 0.013,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5637103982520608,
+      "grad_norm": 0.8453183610999317,
+      "learning_rate": 4.214967684269495e-06,
+      "loss": 0.0069,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5639090277088092,
+      "grad_norm": 0.3829389981415355,
+      "learning_rate": 4.211790225883372e-06,
+      "loss": 0.007,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5641076571655577,
+      "grad_norm": 0.45669163404263635,
+      "learning_rate": 4.208613093894126e-06,
+      "loss": 0.0141,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5643062866223061,
+      "grad_norm": 0.833349864065106,
+      "learning_rate": 4.205436289617406e-06,
+      "loss": 0.0169,
+      "step": 2841
+    },
+    {
+      "epoch": 0.5645049160790545,
+      "grad_norm": 0.6077279896370484,
+      "learning_rate": 4.2022598143687224e-06,
+      "loss": 0.0123,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5647035455358029,
+      "grad_norm": 0.38738975198937353,
+      "learning_rate": 4.199083669463452e-06,
+      "loss": 0.0072,
+      "step": 2843
+    },
+    {
+      "epoch": 0.5649021749925514,
+      "grad_norm": 0.38114824416375975,
+      "learning_rate": 4.195907856216831e-06,
+      "loss": 0.0092,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5651008044492998,
+      "grad_norm": 0.8831215464756551,
+      "learning_rate": 4.192732375943962e-06,
+      "loss": 0.0167,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5652994339060483,
+      "grad_norm": 0.3030833462369814,
+      "learning_rate": 4.189557229959807e-06,
+      "loss": 0.0077,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5654980633627967,
+      "grad_norm": 1.0605500779156578,
+      "learning_rate": 4.186382419579193e-06,
+      "loss": 0.0132,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5656966928195452,
+      "grad_norm": 0.6755075790503954,
+      "learning_rate": 4.183207946116802e-06,
+      "loss": 0.0182,
+      "step": 2848
+    },
+    {
+      "epoch": 0.5658953222762936,
+      "grad_norm": 0.9423178403605319,
+      "learning_rate": 4.180033810887184e-06,
+      "loss": 0.0147,
+      "step": 2849
+    },
+    {
+      "epoch": 0.566093951733042,
+      "grad_norm": 0.6355943716774898,
+      "learning_rate": 4.176860015204743e-06,
+      "loss": 0.0122,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5662925811897904,
+      "grad_norm": 0.46477508165506526,
+      "learning_rate": 4.173686560383745e-06,
+      "loss": 0.0143,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5664912106465388,
+      "grad_norm": 0.489449386116713,
+      "learning_rate": 4.170513447738316e-06,
+      "loss": 0.015,
+      "step": 2852
+    },
+    {
+      "epoch": 0.5666898401032873,
+      "grad_norm": 0.475425921121026,
+      "learning_rate": 4.167340678582437e-06,
+      "loss": 0.0132,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5668884695600358,
+      "grad_norm": 0.5947470913713117,
+      "learning_rate": 4.16416825422995e-06,
+      "loss": 0.0144,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5670870990167842,
+      "grad_norm": 0.8415609819118625,
+      "learning_rate": 4.160996175994551e-06,
+      "loss": 0.0167,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5672857284735326,
+      "grad_norm": 0.9514310803640764,
+      "learning_rate": 4.157824445189796e-06,
+      "loss": 0.0103,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5674843579302811,
+      "grad_norm": 0.6091553295909007,
+      "learning_rate": 4.1546530631290945e-06,
+      "loss": 0.0125,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5676829873870295,
+      "grad_norm": 0.4674705036869269,
+      "learning_rate": 4.151482031125712e-06,
+      "loss": 0.0158,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5678816168437779,
+      "grad_norm": 0.7383391584281234,
+      "learning_rate": 4.148311350492772e-06,
+      "loss": 0.0128,
+      "step": 2859
+    },
+    {
+      "epoch": 0.5680802463005263,
+      "grad_norm": 1.9683596728322885,
+      "learning_rate": 4.145141022543248e-06,
+      "loss": 0.0126,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5682788757572748,
+      "grad_norm": 0.717958085427442,
+      "learning_rate": 4.1419710485899715e-06,
+      "loss": 0.0106,
+      "step": 2861
+    },
+    {
+      "epoch": 0.5684775052140233,
+      "grad_norm": 0.7306809802401357,
+      "learning_rate": 4.138801429945624e-06,
+      "loss": 0.0088,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5686761346707717,
+      "grad_norm": 0.3406143943303101,
+      "learning_rate": 4.135632167922742e-06,
+      "loss": 0.0104,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5688747641275201,
+      "grad_norm": 0.35075777130300395,
+      "learning_rate": 4.1324632638337134e-06,
+      "loss": 0.0071,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5690733935842686,
+      "grad_norm": 3.6206297212120857,
+      "learning_rate": 4.129294718990779e-06,
+      "loss": 0.0089,
+      "step": 2865
+    },
+    {
+      "epoch": 0.569272023041017,
+      "grad_norm": 0.6092937120736366,
+      "learning_rate": 4.126126534706028e-06,
+      "loss": 0.0129,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5694706524977654,
+      "grad_norm": 0.6599744938071138,
+      "learning_rate": 4.122958712291406e-06,
+      "loss": 0.0132,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5696692819545138,
+      "grad_norm": 0.40817723392650745,
+      "learning_rate": 4.119791253058701e-06,
+      "loss": 0.0155,
+      "step": 2868
+    },
+    {
+      "epoch": 0.5698679114112623,
+      "grad_norm": 0.7207327641411628,
+      "learning_rate": 4.1166241583195596e-06,
+      "loss": 0.0149,
+      "step": 2869
+    },
+    {
+      "epoch": 0.5700665408680107,
+      "grad_norm": 0.5234949232683128,
+      "learning_rate": 4.113457429385468e-06,
+      "loss": 0.0089,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5702651703247592,
+      "grad_norm": 0.5297434498649813,
+      "learning_rate": 4.110291067567766e-06,
+      "loss": 0.0107,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5704637997815076,
+      "grad_norm": 0.5678519577896356,
+      "learning_rate": 4.107125074177643e-06,
+      "loss": 0.0092,
+      "step": 2872
+    },
+    {
+      "epoch": 0.570662429238256,
+      "grad_norm": 0.35857068739907116,
+      "learning_rate": 4.103959450526133e-06,
+      "loss": 0.0089,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5708610586950045,
+      "grad_norm": 0.7083827991263594,
+      "learning_rate": 4.100794197924117e-06,
+      "loss": 0.0118,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5710596881517529,
+      "grad_norm": 0.5189059502478018,
+      "learning_rate": 4.097629317682322e-06,
+      "loss": 0.008,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5712583176085013,
+      "grad_norm": 1.0428771302955275,
+      "learning_rate": 4.094464811111323e-06,
+      "loss": 0.0139,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5714569470652497,
+      "grad_norm": 0.711904256313604,
+      "learning_rate": 4.091300679521539e-06,
+      "loss": 0.0162,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5716555765219982,
+      "grad_norm": 0.32525294522509485,
+      "learning_rate": 4.088136924223235e-06,
+      "loss": 0.0106,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5718542059787467,
+      "grad_norm": 0.6302841695249592,
+      "learning_rate": 4.084973546526517e-06,
+      "loss": 0.0115,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5720528354354951,
+      "grad_norm": 0.828179414805218,
+      "learning_rate": 4.081810547741336e-06,
+      "loss": 0.0161,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5722514648922435,
+      "grad_norm": 0.4625165253053684,
+      "learning_rate": 4.078647929177489e-06,
+      "loss": 0.0064,
+      "step": 2881
+    },
+    {
+      "epoch": 0.572450094348992,
+      "grad_norm": 0.6812644619662183,
+      "learning_rate": 4.075485692144611e-06,
+      "loss": 0.0173,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5726487238057404,
+      "grad_norm": 0.8313655313026633,
+      "learning_rate": 4.072323837952181e-06,
+      "loss": 0.0144,
+      "step": 2883
+    },
+    {
+      "epoch": 0.5728473532624888,
+      "grad_norm": 0.45743710231818924,
+      "learning_rate": 4.069162367909522e-06,
+      "loss": 0.0113,
+      "step": 2884
+    },
+    {
+      "epoch": 0.5730459827192372,
+      "grad_norm": 0.9030179886207947,
+      "learning_rate": 4.0660012833257945e-06,
+      "loss": 0.0153,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5732446121759857,
+      "grad_norm": 0.7782720108694118,
+      "learning_rate": 4.062840585510001e-06,
+      "loss": 0.01,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5734432416327341,
+      "grad_norm": 0.40236050134983353,
+      "learning_rate": 4.05968027577098e-06,
+      "loss": 0.0071,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5736418710894826,
+      "grad_norm": 0.6486277794089854,
+      "learning_rate": 4.056520355417418e-06,
+      "loss": 0.01,
+      "step": 2888
+    },
+    {
+      "epoch": 0.573840500546231,
+      "grad_norm": 0.5243182245221714,
+      "learning_rate": 4.053360825757831e-06,
+      "loss": 0.0125,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5740391300029795,
+      "grad_norm": 0.6340711859635116,
+      "learning_rate": 4.050201688100577e-06,
+      "loss": 0.0114,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5742377594597279,
+      "grad_norm": 0.44387658891634135,
+      "learning_rate": 4.047042943753853e-06,
+      "loss": 0.0091,
+      "step": 2891
+    },
+    {
+      "epoch": 0.5744363889164763,
+      "grad_norm": 0.2538412173316712,
+      "learning_rate": 4.043884594025692e-06,
+      "loss": 0.0054,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5746350183732247,
+      "grad_norm": 0.6434174001298162,
+      "learning_rate": 4.040726640223967e-06,
+      "loss": 0.0163,
+      "step": 2893
+    },
+    {
+      "epoch": 0.5748336478299731,
+      "grad_norm": 0.4516311123319802,
+      "learning_rate": 4.037569083656374e-06,
+      "loss": 0.0088,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5750322772867216,
+      "grad_norm": 0.6032522782543573,
+      "learning_rate": 4.034411925630462e-06,
+      "loss": 0.0184,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5752309067434701,
+      "grad_norm": 0.5036736272759907,
+      "learning_rate": 4.031255167453604e-06,
+      "loss": 0.006,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5754295362002185,
+      "grad_norm": 0.5616796264869481,
+      "learning_rate": 4.028098810433012e-06,
+      "loss": 0.02,
+      "step": 2897
+    },
+    {
+      "epoch": 0.575628165656967,
+      "grad_norm": 0.7195746914339065,
+      "learning_rate": 4.024942855875728e-06,
+      "loss": 0.0204,
+      "step": 2898
+    },
+    {
+      "epoch": 0.5758267951137154,
+      "grad_norm": 0.681543673069867,
+      "learning_rate": 4.021787305088633e-06,
+      "loss": 0.0143,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5760254245704638,
+      "grad_norm": 0.4669150650713162,
+      "learning_rate": 4.0186321593784325e-06,
+      "loss": 0.0089,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5762240540272122,
+      "grad_norm": 0.33038421962080927,
+      "learning_rate": 4.015477420051673e-06,
+      "loss": 0.0096,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5764226834839606,
+      "grad_norm": 0.3634840974532087,
+      "learning_rate": 4.012323088414729e-06,
+      "loss": 0.0104,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5766213129407091,
+      "grad_norm": 0.8343922468702314,
+      "learning_rate": 4.009169165773804e-06,
+      "loss": 0.0117,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5768199423974576,
+      "grad_norm": 0.6495727500722914,
+      "learning_rate": 4.0060156534349355e-06,
+      "loss": 0.0125,
+      "step": 2904
+    },
+    {
+      "epoch": 0.577018571854206,
+      "grad_norm": 0.6135698010073438,
+      "learning_rate": 4.00286255270399e-06,
+      "loss": 0.0146,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5772172013109544,
+      "grad_norm": 0.5573911672484038,
+      "learning_rate": 3.9997098648866624e-06,
+      "loss": 0.0124,
+      "step": 2906
+    },
+    {
+      "epoch": 0.5774158307677029,
+      "grad_norm": 0.40620283454662026,
+      "learning_rate": 3.996557591288477e-06,
+      "loss": 0.0106,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5776144602244513,
+      "grad_norm": 0.7192076124521594,
+      "learning_rate": 3.99340573321479e-06,
+      "loss": 0.0102,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5778130896811997,
+      "grad_norm": 0.4360510467100798,
+      "learning_rate": 3.99025429197078e-06,
+      "loss": 0.0141,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5780117191379481,
+      "grad_norm": 0.463212787813459,
+      "learning_rate": 3.987103268861457e-06,
+      "loss": 0.0111,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5782103485946966,
+      "grad_norm": 0.8251451590325685,
+      "learning_rate": 3.983952665191656e-06,
+      "loss": 0.0142,
+      "step": 2911
+    },
+    {
+      "epoch": 0.578408978051445,
+      "grad_norm": 0.761966750523954,
+      "learning_rate": 3.980802482266038e-06,
+      "loss": 0.0163,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5786076075081935,
+      "grad_norm": 0.6924212159837746,
+      "learning_rate": 3.977652721389092e-06,
+      "loss": 0.0099,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5788062369649419,
+      "grad_norm": 0.3979725594269708,
+      "learning_rate": 3.97450338386513e-06,
+      "loss": 0.0085,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5790048664216904,
+      "grad_norm": 0.5618585012904748,
+      "learning_rate": 3.97135447099829e-06,
+      "loss": 0.0127,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5792034958784388,
+      "grad_norm": 0.7497853838343431,
+      "learning_rate": 3.968205984092533e-06,
+      "loss": 0.0108,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5794021253351872,
+      "grad_norm": 0.5120214709576271,
+      "learning_rate": 3.965057924451648e-06,
+      "loss": 0.0131,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5796007547919356,
+      "grad_norm": 0.6871961008819738,
+      "learning_rate": 3.961910293379236e-06,
+      "loss": 0.0144,
+      "step": 2918
+    },
+    {
+      "epoch": 0.579799384248684,
+      "grad_norm": 0.41734721830693366,
+      "learning_rate": 3.958763092178734e-06,
+      "loss": 0.01,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5799980137054325,
+      "grad_norm": 0.4036222914847742,
+      "learning_rate": 3.955616322153391e-06,
+      "loss": 0.0105,
+      "step": 2920
+    },
+    {
+      "epoch": 0.580196643162181,
+      "grad_norm": 0.6130853508919999,
+      "learning_rate": 3.952469984606285e-06,
+      "loss": 0.0134,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5803952726189294,
+      "grad_norm": 0.3509331583951365,
+      "learning_rate": 3.949324080840309e-06,
+      "loss": 0.0098,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5805939020756778,
+      "grad_norm": 0.37833538757507784,
+      "learning_rate": 3.946178612158178e-06,
+      "loss": 0.0134,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5807925315324263,
+      "grad_norm": 0.46980032869709154,
+      "learning_rate": 3.94303357986243e-06,
+      "loss": 0.0093,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5809911609891747,
+      "grad_norm": 0.47678450473727596,
+      "learning_rate": 3.939888985255415e-06,
+      "loss": 0.0156,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5811897904459231,
+      "grad_norm": 0.48897265097095066,
+      "learning_rate": 3.9367448296393115e-06,
+      "loss": 0.0082,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5813884199026715,
+      "grad_norm": 0.39022191785178684,
+      "learning_rate": 3.93360111431611e-06,
+      "loss": 0.0097,
+      "step": 2927
+    },
+    {
+      "epoch": 0.58158704935942,
+      "grad_norm": 0.9510163155213661,
+      "learning_rate": 3.930457840587618e-06,
+      "loss": 0.0189,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5817856788161684,
+      "grad_norm": 0.4727876728395284,
+      "learning_rate": 3.927315009755464e-06,
+      "loss": 0.0142,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5819843082729169,
+      "grad_norm": 0.7381075193685331,
+      "learning_rate": 3.92417262312109e-06,
+      "loss": 0.0111,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5821829377296653,
+      "grad_norm": 0.6627709815131074,
+      "learning_rate": 3.921030681985755e-06,
+      "loss": 0.0108,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5823815671864138,
+      "grad_norm": 0.49816760107685104,
+      "learning_rate": 3.917889187650533e-06,
+      "loss": 0.012,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5825801966431622,
+      "grad_norm": 0.7768642446479597,
+      "learning_rate": 3.914748141416317e-06,
+      "loss": 0.0104,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5827788260999106,
+      "grad_norm": 0.2884247629212007,
+      "learning_rate": 3.9116075445838075e-06,
+      "loss": 0.0094,
+      "step": 2934
+    },
+    {
+      "epoch": 0.582977455556659,
+      "grad_norm": 0.47011413952328,
+      "learning_rate": 3.908467398453524e-06,
+      "loss": 0.0066,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5831760850134075,
+      "grad_norm": 0.4443819609591886,
+      "learning_rate": 3.905327704325799e-06,
+      "loss": 0.0104,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5833747144701559,
+      "grad_norm": 0.3839458518134734,
+      "learning_rate": 3.902188463500774e-06,
+      "loss": 0.0119,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5835733439269044,
+      "grad_norm": 0.6111807494928548,
+      "learning_rate": 3.899049677278407e-06,
+      "loss": 0.013,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5837719733836528,
+      "grad_norm": 0.60033600270167,
+      "learning_rate": 3.895911346958466e-06,
+      "loss": 0.013,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5839706028404013,
+      "grad_norm": 0.2970921298615577,
+      "learning_rate": 3.892773473840531e-06,
+      "loss": 0.0074,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5841692322971497,
+      "grad_norm": 0.610226222580878,
+      "learning_rate": 3.889636059223993e-06,
+      "loss": 0.0167,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5843678617538981,
+      "grad_norm": 0.6862312895558023,
+      "learning_rate": 3.886499104408051e-06,
+      "loss": 0.0121,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5845664912106465,
+      "grad_norm": 0.4404684118718429,
+      "learning_rate": 3.883362610691711e-06,
+      "loss": 0.0142,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5847651206673949,
+      "grad_norm": 0.9493244733608,
+      "learning_rate": 3.880226579373799e-06,
+      "loss": 0.0111,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5849637501241434,
+      "grad_norm": 0.7593496884130583,
+      "learning_rate": 3.877091011752938e-06,
+      "loss": 0.015,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5851623795808919,
+      "grad_norm": 0.6721422032614358,
+      "learning_rate": 3.8739559091275646e-06,
+      "loss": 0.0139,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5853610090376403,
+      "grad_norm": 0.48578862136014994,
+      "learning_rate": 3.870821272795922e-06,
+      "loss": 0.0087,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5855596384943887,
+      "grad_norm": 0.3783706141485433,
+      "learning_rate": 3.867687104056059e-06,
+      "loss": 0.0068,
+      "step": 2948
+    },
+    {
+      "epoch": 0.5857582679511372,
+      "grad_norm": 0.7342026067918515,
+      "learning_rate": 3.864553404205833e-06,
+      "loss": 0.0154,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5859568974078856,
+      "grad_norm": 0.45238663502264936,
+      "learning_rate": 3.861420174542903e-06,
+      "loss": 0.0132,
+      "step": 2950
+    },
+    {
+      "epoch": 0.586155526864634,
+      "grad_norm": 0.5760123588598605,
+      "learning_rate": 3.85828741636474e-06,
+      "loss": 0.0088,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5863541563213824,
+      "grad_norm": 0.37486015440914744,
+      "learning_rate": 3.855155130968616e-06,
+      "loss": 0.0137,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5865527857781309,
+      "grad_norm": 0.4390439990790977,
+      "learning_rate": 3.852023319651605e-06,
+      "loss": 0.011,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5867514152348793,
+      "grad_norm": 0.9686179462089476,
+      "learning_rate": 3.848891983710587e-06,
+      "loss": 0.0108,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5869500446916278,
+      "grad_norm": 0.7875155365033674,
+      "learning_rate": 3.845761124442246e-06,
+      "loss": 0.0238,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5871486741483762,
+      "grad_norm": 0.38882678317489044,
+      "learning_rate": 3.842630743143068e-06,
+      "loss": 0.0119,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5873473036051247,
+      "grad_norm": 0.6336228129818564,
+      "learning_rate": 3.839500841109338e-06,
+      "loss": 0.0104,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5875459330618731,
+      "grad_norm": 0.5805224655559802,
+      "learning_rate": 3.836371419637149e-06,
+      "loss": 0.014,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5877445625186215,
+      "grad_norm": 0.27676008745737823,
+      "learning_rate": 3.833242480022391e-06,
+      "loss": 0.0075,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5879431919753699,
+      "grad_norm": 0.30567765332023794,
+      "learning_rate": 3.8301140235607525e-06,
+      "loss": 0.0074,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5881418214321184,
+      "grad_norm": 0.3651536661757168,
+      "learning_rate": 3.826986051547726e-06,
+      "loss": 0.008,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5883404508888668,
+      "grad_norm": 0.5291250242062171,
+      "learning_rate": 3.8238585652786004e-06,
+      "loss": 0.0109,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5885390803456153,
+      "grad_norm": 0.3137663400127696,
+      "learning_rate": 3.820731566048466e-06,
+      "loss": 0.0087,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5887377098023637,
+      "grad_norm": 2.235523320234743,
+      "learning_rate": 3.817605055152208e-06,
+      "loss": 0.012,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5889363392591122,
+      "grad_norm": 0.3358171761925821,
+      "learning_rate": 3.814479033884514e-06,
+      "loss": 0.008,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5891349687158606,
+      "grad_norm": 0.987905521258979,
+      "learning_rate": 3.8113535035398637e-06,
+      "loss": 0.0168,
+      "step": 2966
+    },
+    {
+      "epoch": 0.589333598172609,
+      "grad_norm": 0.6892134567235605,
+      "learning_rate": 3.8082284654125373e-06,
+      "loss": 0.0164,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5895322276293574,
+      "grad_norm": 0.987943395384034,
+      "learning_rate": 3.805103920796609e-06,
+      "loss": 0.0133,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5897308570861058,
+      "grad_norm": 0.8509307481721003,
+      "learning_rate": 3.8019798709859512e-06,
+      "loss": 0.0116,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5899294865428543,
+      "grad_norm": 0.759142605103621,
+      "learning_rate": 3.79885631727423e-06,
+      "loss": 0.0115,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5901281159996027,
+      "grad_norm": 0.24316763299724442,
+      "learning_rate": 3.7957332609549037e-06,
+      "loss": 0.0069,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5903267454563512,
+      "grad_norm": 0.4485372299375667,
+      "learning_rate": 3.792610703321229e-06,
+      "loss": 0.0086,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5905253749130996,
+      "grad_norm": 0.588418063546346,
+      "learning_rate": 3.789488645666253e-06,
+      "loss": 0.0127,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5907240043698481,
+      "grad_norm": 0.36059543710830055,
+      "learning_rate": 3.7863670892828156e-06,
+      "loss": 0.012,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5909226338265965,
+      "grad_norm": 0.902669268812175,
+      "learning_rate": 3.783246035463551e-06,
+      "loss": 0.014,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5911212632833449,
+      "grad_norm": 0.31958820455962267,
+      "learning_rate": 3.780125485500885e-06,
+      "loss": 0.0087,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5913198927400933,
+      "grad_norm": 0.4986550681651737,
+      "learning_rate": 3.777005440687035e-06,
+      "loss": 0.0098,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5915185221968418,
+      "grad_norm": 0.7580605603680164,
+      "learning_rate": 3.773885902314006e-06,
+      "loss": 0.0241,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5917171516535902,
+      "grad_norm": 0.2330131586895299,
+      "learning_rate": 3.770766871673598e-06,
+      "loss": 0.0069,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5919157811103387,
+      "grad_norm": 0.2919054898745666,
+      "learning_rate": 3.7676483500573966e-06,
+      "loss": 0.0046,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5921144105670871,
+      "grad_norm": 0.38396343904172237,
+      "learning_rate": 3.76453033875678e-06,
+      "loss": 0.0065,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5923130400238356,
+      "grad_norm": 0.8594455212714647,
+      "learning_rate": 3.761412839062911e-06,
+      "loss": 0.0133,
+      "step": 2982
+    },
+    {
+      "epoch": 0.592511669480584,
+      "grad_norm": 0.20526215304091466,
+      "learning_rate": 3.7582958522667466e-06,
+      "loss": 0.0043,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5927102989373324,
+      "grad_norm": 0.3004063766309997,
+      "learning_rate": 3.7551793796590263e-06,
+      "loss": 0.0062,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5929089283940808,
+      "grad_norm": 0.41297527826978053,
+      "learning_rate": 3.7520634225302788e-06,
+      "loss": 0.0105,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5931075578508292,
+      "grad_norm": 0.8190693978061193,
+      "learning_rate": 3.7489479821708173e-06,
+      "loss": 0.0171,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5933061873075777,
+      "grad_norm": 0.4697013617601209,
+      "learning_rate": 3.7458330598707443e-06,
+      "loss": 0.0065,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5935048167643262,
+      "grad_norm": 0.5384383900255248,
+      "learning_rate": 3.7427186569199456e-06,
+      "loss": 0.0128,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5937034462210746,
+      "grad_norm": 0.753459781552131,
+      "learning_rate": 3.739604774608092e-06,
+      "loss": 0.02,
+      "step": 2989
+    },
+    {
+      "epoch": 0.593902075677823,
+      "grad_norm": 0.49028554186530215,
+      "learning_rate": 3.7364914142246383e-06,
+      "loss": 0.0101,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5941007051345715,
+      "grad_norm": 0.7809148931549772,
+      "learning_rate": 3.733378577058825e-06,
+      "loss": 0.0146,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5942993345913199,
+      "grad_norm": 0.7722259717302931,
+      "learning_rate": 3.7302662643996747e-06,
+      "loss": 0.0141,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5944979640480683,
+      "grad_norm": 0.6140405841322572,
+      "learning_rate": 3.7271544775359906e-06,
+      "loss": 0.0162,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5946965935048167,
+      "grad_norm": 0.7586513477389154,
+      "learning_rate": 3.7240432177563646e-06,
+      "loss": 0.0132,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5948952229615652,
+      "grad_norm": 0.6630658842640412,
+      "learning_rate": 3.720932486349165e-06,
+      "loss": 0.0153,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5950938524183136,
+      "grad_norm": 0.6232367511499796,
+      "learning_rate": 3.7178222846025404e-06,
+      "loss": 0.0097,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5952924818750621,
+      "grad_norm": 0.38137155843187615,
+      "learning_rate": 3.7147126138044243e-06,
+      "loss": 0.0072,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5954911113318105,
+      "grad_norm": 0.6383755710136322,
+      "learning_rate": 3.7116034752425277e-06,
+      "loss": 0.0124,
+      "step": 2998
+    },
+    {
+      "epoch": 0.595689740788559,
+      "grad_norm": 0.5596917796188703,
+      "learning_rate": 3.708494870204342e-06,
+      "loss": 0.0158,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5958883702453074,
+      "grad_norm": 0.42835170745607165,
+      "learning_rate": 3.7053867999771366e-06,
+      "loss": 0.0125,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5960869997020558,
+      "grad_norm": 0.39430591078922234,
+      "learning_rate": 3.702279265847961e-06,
+      "loss": 0.0099,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5962856291588042,
+      "grad_norm": 0.509207291742738,
+      "learning_rate": 3.6991722691036423e-06,
+      "loss": 0.0104,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5964842586155527,
+      "grad_norm": 0.8424408408284407,
+      "learning_rate": 3.6960658110307844e-06,
+      "loss": 0.0109,
+      "step": 3003
+    },
+    {
+      "epoch": 0.5966828880723011,
+      "grad_norm": 0.7528576265240631,
+      "learning_rate": 3.6929598929157682e-06,
+      "loss": 0.0151,
+      "step": 3004
+    },
+    {
+      "epoch": 0.5968815175290496,
+      "grad_norm": 0.45476504766802717,
+      "learning_rate": 3.689854516044752e-06,
+      "loss": 0.0143,
+      "step": 3005
+    },
+    {
+      "epoch": 0.597080146985798,
+      "grad_norm": 0.44031126237277396,
+      "learning_rate": 3.6867496817036674e-06,
+      "loss": 0.0138,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5972787764425465,
+      "grad_norm": 0.5965406726879514,
+      "learning_rate": 3.6836453911782244e-06,
+      "loss": 0.0097,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5974774058992949,
+      "grad_norm": 0.32200058430663714,
+      "learning_rate": 3.680541645753908e-06,
+      "loss": 0.0065,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5976760353560433,
+      "grad_norm": 0.35025081668798586,
+      "learning_rate": 3.677438446715974e-06,
+      "loss": 0.0148,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5978746648127917,
+      "grad_norm": 0.732734140663941,
+      "learning_rate": 3.6743357953494554e-06,
+      "loss": 0.0232,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5980732942695401,
+      "grad_norm": 0.9147359044098355,
+      "learning_rate": 3.6712336929391558e-06,
+      "loss": 0.0171,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5982719237262886,
+      "grad_norm": 0.47195978016132334,
+      "learning_rate": 3.6681321407696546e-06,
+      "loss": 0.0108,
+      "step": 3012
+    },
+    {
+      "epoch": 0.598470553183037,
+      "grad_norm": 0.3255368283156676,
+      "learning_rate": 3.665031140125299e-06,
+      "loss": 0.0082,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5986691826397855,
+      "grad_norm": 0.4043095691857067,
+      "learning_rate": 3.661930692290211e-06,
+      "loss": 0.0111,
+      "step": 3014
+    },
+    {
+      "epoch": 0.598867812096534,
+      "grad_norm": 0.8606549728317674,
+      "learning_rate": 3.658830798548284e-06,
+      "loss": 0.0208,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5990664415532824,
+      "grad_norm": 0.31791862321187564,
+      "learning_rate": 3.6557314601831804e-06,
+      "loss": 0.0075,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5992650710100308,
+      "grad_norm": 0.61615355943961,
+      "learning_rate": 3.6526326784783328e-06,
+      "loss": 0.0188,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5994637004667792,
+      "grad_norm": 0.3392910567038716,
+      "learning_rate": 3.649534454716942e-06,
+      "loss": 0.0071,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5996623299235276,
+      "grad_norm": 0.3407654018111544,
+      "learning_rate": 3.646436790181983e-06,
+      "loss": 0.0091,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5998609593802761,
+      "grad_norm": 0.43481756324822807,
+      "learning_rate": 3.643339686156193e-06,
+      "loss": 0.007,
+      "step": 3020
+    },
+    {
+      "epoch": 0.6000595888370245,
+      "grad_norm": 0.42659281115467734,
+      "learning_rate": 3.6402431439220807e-06,
+      "loss": 0.0081,
+      "step": 3021
+    },
+    {
+      "epoch": 0.600258218293773,
+      "grad_norm": 0.7240516093008215,
+      "learning_rate": 3.6371471647619212e-06,
+      "loss": 0.0172,
+      "step": 3022
+    },
+    {
+      "epoch": 0.6004568477505214,
+      "grad_norm": 0.2238508808138508,
+      "learning_rate": 3.6340517499577552e-06,
+      "loss": 0.0068,
+      "step": 3023
+    },
+    {
+      "epoch": 0.6006554772072699,
+      "grad_norm": 0.641644301510662,
+      "learning_rate": 3.6309569007913926e-06,
+      "loss": 0.0192,
+      "step": 3024
+    },
+    {
+      "epoch": 0.6008541066640183,
+      "grad_norm": 0.6086067117270952,
+      "learning_rate": 3.6278626185444043e-06,
+      "loss": 0.0146,
+      "step": 3025
+    },
+    {
+      "epoch": 0.6010527361207667,
+      "grad_norm": 0.562268484312575,
+      "learning_rate": 3.624768904498133e-06,
+      "loss": 0.0127,
+      "step": 3026
+    },
+    {
+      "epoch": 0.6012513655775151,
+      "grad_norm": 0.5503772462099394,
+      "learning_rate": 3.6216757599336817e-06,
+      "loss": 0.0122,
+      "step": 3027
+    },
+    {
+      "epoch": 0.6014499950342636,
+      "grad_norm": 0.3503335156258801,
+      "learning_rate": 3.6185831861319175e-06,
+      "loss": 0.0096,
+      "step": 3028
+    },
+    {
+      "epoch": 0.601648624491012,
+      "grad_norm": 0.4135501081556177,
+      "learning_rate": 3.6154911843734726e-06,
+      "loss": 0.0098,
+      "step": 3029
+    },
+    {
+      "epoch": 0.6018472539477605,
+      "grad_norm": 0.3319077445993011,
+      "learning_rate": 3.612399755938741e-06,
+      "loss": 0.0095,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6020458834045089,
+      "grad_norm": 0.4357565471543654,
+      "learning_rate": 3.609308902107882e-06,
+      "loss": 0.0093,
+      "step": 3031
+    },
+    {
+      "epoch": 0.6022445128612574,
+      "grad_norm": 0.3754875478839089,
+      "learning_rate": 3.6062186241608127e-06,
+      "loss": 0.0067,
+      "step": 3032
+    },
+    {
+      "epoch": 0.6024431423180058,
+      "grad_norm": 0.4655009047495604,
+      "learning_rate": 3.603128923377216e-06,
+      "loss": 0.0104,
+      "step": 3033
+    },
+    {
+      "epoch": 0.6026417717747542,
+      "grad_norm": 0.37526412933260067,
+      "learning_rate": 3.6000398010365335e-06,
+      "loss": 0.0102,
+      "step": 3034
+    },
+    {
+      "epoch": 0.6028404012315026,
+      "grad_norm": 0.37466649162476384,
+      "learning_rate": 3.5969512584179676e-06,
+      "loss": 0.0062,
+      "step": 3035
+    },
+    {
+      "epoch": 0.603039030688251,
+      "grad_norm": 0.6376237051267966,
+      "learning_rate": 3.5938632968004816e-06,
+      "loss": 0.007,
+      "step": 3036
+    },
+    {
+      "epoch": 0.6032376601449995,
+      "grad_norm": 0.5472784553576067,
+      "learning_rate": 3.590775917462795e-06,
+      "loss": 0.0087,
+      "step": 3037
+    },
+    {
+      "epoch": 0.6034362896017479,
+      "grad_norm": 0.35009623082706764,
+      "learning_rate": 3.5876891216833898e-06,
+      "loss": 0.0111,
+      "step": 3038
+    },
+    {
+      "epoch": 0.6036349190584964,
+      "grad_norm": 0.46974213511879326,
+      "learning_rate": 3.5846029107405043e-06,
+      "loss": 0.0095,
+      "step": 3039
+    },
+    {
+      "epoch": 0.6038335485152448,
+      "grad_norm": 0.7485818959883949,
+      "learning_rate": 3.581517285912137e-06,
+      "loss": 0.0173,
+      "step": 3040
+    },
+    {
+      "epoch": 0.6040321779719933,
+      "grad_norm": 0.38815346960654723,
+      "learning_rate": 3.578432248476041e-06,
+      "loss": 0.0086,
+      "step": 3041
+    },
+    {
+      "epoch": 0.6042308074287417,
+      "grad_norm": 0.5263522252501535,
+      "learning_rate": 3.5753477997097274e-06,
+      "loss": 0.0089,
+      "step": 3042
+    },
+    {
+      "epoch": 0.6044294368854901,
+      "grad_norm": 0.5860186189203216,
+      "learning_rate": 3.5722639408904613e-06,
+      "loss": 0.0121,
+      "step": 3043
+    },
+    {
+      "epoch": 0.6046280663422385,
+      "grad_norm": 0.5281102457455624,
+      "learning_rate": 3.5691806732952695e-06,
+      "loss": 0.0121,
+      "step": 3044
+    },
+    {
+      "epoch": 0.604826695798987,
+      "grad_norm": 0.6500090193718753,
+      "learning_rate": 3.5660979982009283e-06,
+      "loss": 0.0136,
+      "step": 3045
+    },
+    {
+      "epoch": 0.6050253252557354,
+      "grad_norm": 0.39939760825863435,
+      "learning_rate": 3.563015916883969e-06,
+      "loss": 0.0085,
+      "step": 3046
+    },
+    {
+      "epoch": 0.6052239547124839,
+      "grad_norm": 0.347841392980447,
+      "learning_rate": 3.5599344306206797e-06,
+      "loss": 0.0106,
+      "step": 3047
+    },
+    {
+      "epoch": 0.6054225841692323,
+      "grad_norm": 0.9314772406942418,
+      "learning_rate": 3.5568535406871006e-06,
+      "loss": 0.0122,
+      "step": 3048
+    },
+    {
+      "epoch": 0.6056212136259808,
+      "grad_norm": 0.4754175168943589,
+      "learning_rate": 3.553773248359026e-06,
+      "loss": 0.0114,
+      "step": 3049
+    },
+    {
+      "epoch": 0.6058198430827292,
+      "grad_norm": 0.3708403797175746,
+      "learning_rate": 3.5506935549119994e-06,
+      "loss": 0.0091,
+      "step": 3050
+    },
+    {
+      "epoch": 0.6060184725394776,
+      "grad_norm": 1.7346629552296347,
+      "learning_rate": 3.547614461621321e-06,
+      "loss": 0.013,
+      "step": 3051
+    },
+    {
+      "epoch": 0.606217101996226,
+      "grad_norm": 0.40695318664902086,
+      "learning_rate": 3.5445359697620396e-06,
+      "loss": 0.0092,
+      "step": 3052
+    },
+    {
+      "epoch": 0.6064157314529744,
+      "grad_norm": 0.7125792193716718,
+      "learning_rate": 3.541458080608956e-06,
+      "loss": 0.0119,
+      "step": 3053
+    },
+    {
+      "epoch": 0.6066143609097229,
+      "grad_norm": 0.6915939424187886,
+      "learning_rate": 3.5383807954366207e-06,
+      "loss": 0.0168,
+      "step": 3054
+    },
+    {
+      "epoch": 0.6068129903664713,
+      "grad_norm": 1.730306247282683,
+      "learning_rate": 3.5353041155193333e-06,
+      "loss": 0.0073,
+      "step": 3055
+    },
+    {
+      "epoch": 0.6070116198232198,
+      "grad_norm": 0.438350129352838,
+      "learning_rate": 3.5322280421311462e-06,
+      "loss": 0.0095,
+      "step": 3056
+    },
+    {
+      "epoch": 0.6072102492799683,
+      "grad_norm": 1.0041964816668,
+      "learning_rate": 3.5291525765458555e-06,
+      "loss": 0.0138,
+      "step": 3057
+    },
+    {
+      "epoch": 0.6074088787367167,
+      "grad_norm": 0.48577362615681124,
+      "learning_rate": 3.5260777200370108e-06,
+      "loss": 0.0109,
+      "step": 3058
+    },
+    {
+      "epoch": 0.6076075081934651,
+      "grad_norm": 0.5217130396983579,
+      "learning_rate": 3.5230034738779062e-06,
+      "loss": 0.0166,
+      "step": 3059
+    },
+    {
+      "epoch": 0.6078061376502135,
+      "grad_norm": 0.519867329149172,
+      "learning_rate": 3.519929839341586e-06,
+      "loss": 0.0135,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6080047671069619,
+      "grad_norm": 0.5343544411448541,
+      "learning_rate": 3.5168568177008343e-06,
+      "loss": 0.0121,
+      "step": 3061
+    },
+    {
+      "epoch": 0.6082033965637104,
+      "grad_norm": 0.5921739219433378,
+      "learning_rate": 3.51378441022819e-06,
+      "loss": 0.0217,
+      "step": 3062
+    },
+    {
+      "epoch": 0.6084020260204588,
+      "grad_norm": 0.4319692092553288,
+      "learning_rate": 3.5107126181959326e-06,
+      "loss": 0.0087,
+      "step": 3063
+    },
+    {
+      "epoch": 0.6086006554772073,
+      "grad_norm": 0.7628603176542459,
+      "learning_rate": 3.507641442876089e-06,
+      "loss": 0.0079,
+      "step": 3064
+    },
+    {
+      "epoch": 0.6087992849339557,
+      "grad_norm": 0.5049220741692332,
+      "learning_rate": 3.50457088554043e-06,
+      "loss": 0.0047,
+      "step": 3065
+    },
+    {
+      "epoch": 0.6089979143907042,
+      "grad_norm": 1.412081883923703,
+      "learning_rate": 3.5015009474604687e-06,
+      "loss": 0.0155,
+      "step": 3066
+    },
+    {
+      "epoch": 0.6091965438474526,
+      "grad_norm": 1.8218808527341421,
+      "learning_rate": 3.498431629907465e-06,
+      "loss": 0.0156,
+      "step": 3067
+    },
+    {
+      "epoch": 0.609395173304201,
+      "grad_norm": 0.3862970491125843,
+      "learning_rate": 3.4953629341524185e-06,
+      "loss": 0.0083,
+      "step": 3068
+    },
+    {
+      "epoch": 0.6095938027609494,
+      "grad_norm": 0.5034340958370394,
+      "learning_rate": 3.4922948614660755e-06,
+      "loss": 0.0106,
+      "step": 3069
+    },
+    {
+      "epoch": 0.6097924322176979,
+      "grad_norm": 0.6806056800334587,
+      "learning_rate": 3.4892274131189203e-06,
+      "loss": 0.0091,
+      "step": 3070
+    },
+    {
+      "epoch": 0.6099910616744463,
+      "grad_norm": 1.6108884326279738,
+      "learning_rate": 3.4861605903811802e-06,
+      "loss": 0.0169,
+      "step": 3071
+    },
+    {
+      "epoch": 0.6101896911311948,
+      "grad_norm": 0.4012224150673856,
+      "learning_rate": 3.4830943945228243e-06,
+      "loss": 0.0086,
+      "step": 3072
+    },
+    {
+      "epoch": 0.6103883205879432,
+      "grad_norm": 0.5349335803458497,
+      "learning_rate": 3.4800288268135598e-06,
+      "loss": 0.0174,
+      "step": 3073
+    },
+    {
+      "epoch": 0.6105869500446917,
+      "grad_norm": 0.4841940483295753,
+      "learning_rate": 3.4769638885228364e-06,
+      "loss": 0.0168,
+      "step": 3074
+    },
+    {
+      "epoch": 0.6107855795014401,
+      "grad_norm": 0.6441740428833413,
+      "learning_rate": 3.47389958091984e-06,
+      "loss": 0.015,
+      "step": 3075
+    },
+    {
+      "epoch": 0.6109842089581885,
+      "grad_norm": 0.357406903865928,
+      "learning_rate": 3.4708359052735006e-06,
+      "loss": 0.0082,
+      "step": 3076
+    },
+    {
+      "epoch": 0.6111828384149369,
+      "grad_norm": 0.5834140528899641,
+      "learning_rate": 3.4677728628524807e-06,
+      "loss": 0.0115,
+      "step": 3077
+    },
+    {
+      "epoch": 0.6113814678716853,
+      "grad_norm": 0.4928712484829136,
+      "learning_rate": 3.464710454925184e-06,
+      "loss": 0.0132,
+      "step": 3078
+    },
+    {
+      "epoch": 0.6115800973284338,
+      "grad_norm": 0.5979492434923857,
+      "learning_rate": 3.461648682759752e-06,
+      "loss": 0.0095,
+      "step": 3079
+    },
+    {
+      "epoch": 0.6117787267851822,
+      "grad_norm": 0.5154017529353565,
+      "learning_rate": 3.4585875476240584e-06,
+      "loss": 0.0131,
+      "step": 3080
+    },
+    {
+      "epoch": 0.6119773562419307,
+      "grad_norm": 1.054941339463001,
+      "learning_rate": 3.4555270507857174e-06,
+      "loss": 0.0082,
+      "step": 3081
+    },
+    {
+      "epoch": 0.6121759856986791,
+      "grad_norm": 0.7383146639599456,
+      "learning_rate": 3.452467193512078e-06,
+      "loss": 0.0086,
+      "step": 3082
+    },
+    {
+      "epoch": 0.6123746151554276,
+      "grad_norm": 0.46223660532581257,
+      "learning_rate": 3.449407977070225e-06,
+      "loss": 0.008,
+      "step": 3083
+    },
+    {
+      "epoch": 0.612573244612176,
+      "grad_norm": 0.6168524172398883,
+      "learning_rate": 3.4463494027269772e-06,
+      "loss": 0.0111,
+      "step": 3084
+    },
+    {
+      "epoch": 0.6127718740689244,
+      "grad_norm": 0.443610681926009,
+      "learning_rate": 3.443291471748884e-06,
+      "loss": 0.0088,
+      "step": 3085
+    },
+    {
+      "epoch": 0.6129705035256728,
+      "grad_norm": 0.42211292893459496,
+      "learning_rate": 3.4402341854022326e-06,
+      "loss": 0.0115,
+      "step": 3086
+    },
+    {
+      "epoch": 0.6131691329824213,
+      "grad_norm": 0.38959656912953755,
+      "learning_rate": 3.4371775449530444e-06,
+      "loss": 0.0118,
+      "step": 3087
+    },
+    {
+      "epoch": 0.6133677624391697,
+      "grad_norm": 0.482420771604837,
+      "learning_rate": 3.434121551667069e-06,
+      "loss": 0.0124,
+      "step": 3088
+    },
+    {
+      "epoch": 0.6135663918959182,
+      "grad_norm": 0.9713114677647458,
+      "learning_rate": 3.4310662068097915e-06,
+      "loss": 0.0177,
+      "step": 3089
+    },
+    {
+      "epoch": 0.6137650213526666,
+      "grad_norm": 0.41367064378744073,
+      "learning_rate": 3.4280115116464263e-06,
+      "loss": 0.0137,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6139636508094151,
+      "grad_norm": 0.3784820028429733,
+      "learning_rate": 3.4249574674419206e-06,
+      "loss": 0.0163,
+      "step": 3091
+    },
+    {
+      "epoch": 0.6141622802661635,
+      "grad_norm": 0.4813484208222651,
+      "learning_rate": 3.4219040754609497e-06,
+      "loss": 0.0141,
+      "step": 3092
+    },
+    {
+      "epoch": 0.6143609097229119,
+      "grad_norm": 0.28494432612150494,
+      "learning_rate": 3.41885133696792e-06,
+      "loss": 0.0054,
+      "step": 3093
+    },
+    {
+      "epoch": 0.6145595391796603,
+      "grad_norm": 0.5830609541475335,
+      "learning_rate": 3.415799253226969e-06,
+      "loss": 0.0095,
+      "step": 3094
+    },
+    {
+      "epoch": 0.6147581686364088,
+      "grad_norm": 0.32540607111091596,
+      "learning_rate": 3.4127478255019607e-06,
+      "loss": 0.0107,
+      "step": 3095
+    },
+    {
+      "epoch": 0.6149567980931572,
+      "grad_norm": 0.35280809057789947,
+      "learning_rate": 3.409697055056489e-06,
+      "loss": 0.0105,
+      "step": 3096
+    },
+    {
+      "epoch": 0.6151554275499056,
+      "grad_norm": 0.7711115094509354,
+      "learning_rate": 3.406646943153874e-06,
+      "loss": 0.0131,
+      "step": 3097
+    },
+    {
+      "epoch": 0.6153540570066541,
+      "grad_norm": 0.6585705185875225,
+      "learning_rate": 3.4035974910571635e-06,
+      "loss": 0.0089,
+      "step": 3098
+    },
+    {
+      "epoch": 0.6155526864634026,
+      "grad_norm": 0.8014653930270288,
+      "learning_rate": 3.4005487000291336e-06,
+      "loss": 0.0094,
+      "step": 3099
+    },
+    {
+      "epoch": 0.615751315920151,
+      "grad_norm": 0.6004322261525369,
+      "learning_rate": 3.3975005713322852e-06,
+      "loss": 0.0183,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6159499453768994,
+      "grad_norm": 0.5039813081737613,
+      "learning_rate": 3.3944531062288456e-06,
+      "loss": 0.0117,
+      "step": 3101
+    },
+    {
+      "epoch": 0.6161485748336478,
+      "grad_norm": 0.32998773181787877,
+      "learning_rate": 3.391406305980767e-06,
+      "loss": 0.0072,
+      "step": 3102
+    },
+    {
+      "epoch": 0.6163472042903962,
+      "grad_norm": 0.39531334946742824,
+      "learning_rate": 3.388360171849726e-06,
+      "loss": 0.0131,
+      "step": 3103
+    },
+    {
+      "epoch": 0.6165458337471447,
+      "grad_norm": 1.4423036226921424,
+      "learning_rate": 3.3853147050971245e-06,
+      "loss": 0.028,
+      "step": 3104
+    },
+    {
+      "epoch": 0.6167444632038931,
+      "grad_norm": 0.6108617381540014,
+      "learning_rate": 3.382269906984086e-06,
+      "loss": 0.0083,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6169430926606416,
+      "grad_norm": 0.5450037541557885,
+      "learning_rate": 3.3792257787714593e-06,
+      "loss": 0.0192,
+      "step": 3106
+    },
+    {
+      "epoch": 0.61714172211739,
+      "grad_norm": 1.7949375217276946,
+      "learning_rate": 3.376182321719813e-06,
+      "loss": 0.0241,
+      "step": 3107
+    },
+    {
+      "epoch": 0.6173403515741385,
+      "grad_norm": 0.47593830021557054,
+      "learning_rate": 3.3731395370894447e-06,
+      "loss": 0.0137,
+      "step": 3108
+    },
+    {
+      "epoch": 0.6175389810308869,
+      "grad_norm": 0.5799733370382245,
+      "learning_rate": 3.370097426140363e-06,
+      "loss": 0.0119,
+      "step": 3109
+    },
+    {
+      "epoch": 0.6177376104876353,
+      "grad_norm": 0.4307995841914232,
+      "learning_rate": 3.3670559901323054e-06,
+      "loss": 0.0081,
+      "step": 3110
+    },
+    {
+      "epoch": 0.6179362399443837,
+      "grad_norm": 1.0801970697450218,
+      "learning_rate": 3.364015230324725e-06,
+      "loss": 0.0237,
+      "step": 3111
+    },
+    {
+      "epoch": 0.6181348694011322,
+      "grad_norm": 0.4367611467960593,
+      "learning_rate": 3.3609751479768003e-06,
+      "loss": 0.0098,
+      "step": 3112
+    },
+    {
+      "epoch": 0.6183334988578806,
+      "grad_norm": 0.3381363192748383,
+      "learning_rate": 3.3579357443474264e-06,
+      "loss": 0.0083,
+      "step": 3113
+    },
+    {
+      "epoch": 0.6185321283146291,
+      "grad_norm": 0.31029202975354897,
+      "learning_rate": 3.354897020695216e-06,
+      "loss": 0.0069,
+      "step": 3114
+    },
+    {
+      "epoch": 0.6187307577713775,
+      "grad_norm": 0.36141553067284127,
+      "learning_rate": 3.3518589782785016e-06,
+      "loss": 0.0053,
+      "step": 3115
+    },
+    {
+      "epoch": 0.618929387228126,
+      "grad_norm": 0.7205196066259474,
+      "learning_rate": 3.348821618355334e-06,
+      "loss": 0.0174,
+      "step": 3116
+    },
+    {
+      "epoch": 0.6191280166848744,
+      "grad_norm": 0.392173406386398,
+      "learning_rate": 3.345784942183481e-06,
+      "loss": 0.0118,
+      "step": 3117
+    },
+    {
+      "epoch": 0.6193266461416228,
+      "grad_norm": 0.3923813550404335,
+      "learning_rate": 3.342748951020425e-06,
+      "loss": 0.0072,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6195252755983712,
+      "grad_norm": 0.9318581326956561,
+      "learning_rate": 3.3397136461233705e-06,
+      "loss": 0.0158,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6197239050551197,
+      "grad_norm": 0.4305556661534459,
+      "learning_rate": 3.3366790287492323e-06,
+      "loss": 0.01,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6199225345118681,
+      "grad_norm": 0.494409947487882,
+      "learning_rate": 3.3336451001546422e-06,
+      "loss": 0.0114,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6201211639686165,
+      "grad_norm": 0.41455988706859453,
+      "learning_rate": 3.3306118615959483e-06,
+      "loss": 0.006,
+      "step": 3122
+    },
+    {
+      "epoch": 0.620319793425365,
+      "grad_norm": 0.7803252545572082,
+      "learning_rate": 3.32757931432921e-06,
+      "loss": 0.0181,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6205184228821135,
+      "grad_norm": 0.5185588948320111,
+      "learning_rate": 3.3245474596102034e-06,
+      "loss": 0.0094,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6207170523388619,
+      "grad_norm": 0.35619054220848395,
+      "learning_rate": 3.3215162986944145e-06,
+      "loss": 0.0169,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6209156817956103,
+      "grad_norm": 0.6897006764622771,
+      "learning_rate": 3.3184858328370464e-06,
+      "loss": 0.0112,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6211143112523587,
+      "grad_norm": 0.579124162879879,
+      "learning_rate": 3.315456063293011e-06,
+      "loss": 0.0161,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6213129407091071,
+      "grad_norm": 0.45691926633221563,
+      "learning_rate": 3.312426991316933e-06,
+      "loss": 0.0079,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6215115701658556,
+      "grad_norm": 0.3550439857141177,
+      "learning_rate": 3.309398618163148e-06,
+      "loss": 0.01,
+      "step": 3129
+    },
+    {
+      "epoch": 0.621710199622604,
+      "grad_norm": 0.4261114797732587,
+      "learning_rate": 3.306370945085702e-06,
+      "loss": 0.0093,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6219088290793525,
+      "grad_norm": 0.509512632829503,
+      "learning_rate": 3.303343973338352e-06,
+      "loss": 0.0149,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6221074585361009,
+      "grad_norm": 0.33665150006855243,
+      "learning_rate": 3.3003177041745644e-06,
+      "loss": 0.0044,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6223060879928494,
+      "grad_norm": 0.6093443769153506,
+      "learning_rate": 3.297292138847512e-06,
+      "loss": 0.0139,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6225047174495978,
+      "grad_norm": 0.6356168599910234,
+      "learning_rate": 3.2942672786100806e-06,
+      "loss": 0.0119,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6227033469063462,
+      "grad_norm": 0.4716855426406557,
+      "learning_rate": 3.2912431247148603e-06,
+      "loss": 0.0123,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6229019763630946,
+      "grad_norm": 0.26872765852592456,
+      "learning_rate": 3.2882196784141497e-06,
+      "loss": 0.0087,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6231006058198431,
+      "grad_norm": 0.9289996201648486,
+      "learning_rate": 3.285196940959957e-06,
+      "loss": 0.0128,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6232992352765915,
+      "grad_norm": 0.5778170278567689,
+      "learning_rate": 3.2821749136039947e-06,
+      "loss": 0.015,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6234978647333399,
+      "grad_norm": 0.44484550235368014,
+      "learning_rate": 3.2791535975976796e-06,
+      "loss": 0.0112,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6236964941900884,
+      "grad_norm": 0.5650552894350845,
+      "learning_rate": 3.276132994192137e-06,
+      "loss": 0.0168,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6238951236468369,
+      "grad_norm": 0.45586647670638364,
+      "learning_rate": 3.2731131046381946e-06,
+      "loss": 0.0165,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6240937531035853,
+      "grad_norm": 0.3991031664463805,
+      "learning_rate": 3.2700939301863867e-06,
+      "loss": 0.0088,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6242923825603337,
+      "grad_norm": 0.5441221777716391,
+      "learning_rate": 3.2670754720869483e-06,
+      "loss": 0.0091,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6244910120170821,
+      "grad_norm": 0.4508262833394596,
+      "learning_rate": 3.2640577315898232e-06,
+      "loss": 0.0101,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6246896414738305,
+      "grad_norm": 0.5522463595883214,
+      "learning_rate": 3.261040709944653e-06,
+      "loss": 0.0199,
+      "step": 3145
+    },
+    {
+      "epoch": 0.624888270930579,
+      "grad_norm": 0.5241449671995597,
+      "learning_rate": 3.258024408400783e-06,
+      "loss": 0.0155,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6250869003873274,
+      "grad_norm": 0.5548789843023203,
+      "learning_rate": 3.2550088282072614e-06,
+      "loss": 0.0109,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6252855298440759,
+      "grad_norm": 1.0504625440702693,
+      "learning_rate": 3.2519939706128354e-06,
+      "loss": 0.019,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6254841593008243,
+      "grad_norm": 0.4102947730181052,
+      "learning_rate": 3.2489798368659568e-06,
+      "loss": 0.0106,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6256827887575728,
+      "grad_norm": 0.6501870965913413,
+      "learning_rate": 3.245966428214773e-06,
+      "loss": 0.0102,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6258814182143212,
+      "grad_norm": 0.542258565431279,
+      "learning_rate": 3.242953745907136e-06,
+      "loss": 0.0164,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6260800476710696,
+      "grad_norm": 0.43709875345405597,
+      "learning_rate": 3.2399417911905928e-06,
+      "loss": 0.0099,
+      "step": 3152
+    },
+    {
+      "epoch": 0.626278677127818,
+      "grad_norm": 0.4169502805419377,
+      "learning_rate": 3.2369305653123918e-06,
+      "loss": 0.0161,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6264773065845665,
+      "grad_norm": 0.7203211824671325,
+      "learning_rate": 3.2339200695194776e-06,
+      "loss": 0.0132,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6266759360413149,
+      "grad_norm": 0.6337861054133493,
+      "learning_rate": 3.2309103050584943e-06,
+      "loss": 0.0155,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6268745654980633,
+      "grad_norm": 0.5461213073528943,
+      "learning_rate": 3.227901273175783e-06,
+      "loss": 0.018,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6270731949548118,
+      "grad_norm": 0.4288994784155615,
+      "learning_rate": 3.2248929751173796e-06,
+      "loss": 0.013,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6272718244115603,
+      "grad_norm": 0.5336781946873922,
+      "learning_rate": 3.2218854121290167e-06,
+      "loss": 0.0125,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6274704538683087,
+      "grad_norm": 0.39411430837415473,
+      "learning_rate": 3.218878585456124e-06,
+      "loss": 0.0155,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6276690833250571,
+      "grad_norm": 0.7372205200276526,
+      "learning_rate": 3.215872496343826e-06,
+      "loss": 0.0109,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6278677127818055,
+      "grad_norm": 0.31051639973025613,
+      "learning_rate": 3.212867146036939e-06,
+      "loss": 0.0112,
+      "step": 3161
+    },
+    {
+      "epoch": 0.628066342238554,
+      "grad_norm": 0.7517003368889213,
+      "learning_rate": 3.2098625357799777e-06,
+      "loss": 0.0152,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6282649716953024,
+      "grad_norm": 0.4493855979835352,
+      "learning_rate": 3.2068586668171487e-06,
+      "loss": 0.0128,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6284636011520508,
+      "grad_norm": 0.43259693880401673,
+      "learning_rate": 3.2038555403923495e-06,
+      "loss": 0.0105,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6286622306087993,
+      "grad_norm": 0.6757306391799492,
+      "learning_rate": 3.2008531577491726e-06,
+      "loss": 0.015,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6288608600655478,
+      "grad_norm": 0.7200622315806112,
+      "learning_rate": 3.197851520130901e-06,
+      "loss": 0.0127,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6290594895222962,
+      "grad_norm": 0.30945546413393143,
+      "learning_rate": 3.1948506287805105e-06,
+      "loss": 0.0071,
+      "step": 3167
+    },
+    {
+      "epoch": 0.6292581189790446,
+      "grad_norm": 0.5184927164257134,
+      "learning_rate": 3.1918504849406655e-06,
+      "loss": 0.013,
+      "step": 3168
+    },
+    {
+      "epoch": 0.629456748435793,
+      "grad_norm": 0.37944118777802155,
+      "learning_rate": 3.188851089853725e-06,
+      "loss": 0.0121,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6296553778925414,
+      "grad_norm": 0.6108105036736555,
+      "learning_rate": 3.185852444761735e-06,
+      "loss": 0.0175,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6298540073492899,
+      "grad_norm": 0.45049388301001664,
+      "learning_rate": 3.182854550906431e-06,
+      "loss": 0.0119,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6300526368060383,
+      "grad_norm": 0.5933853421541697,
+      "learning_rate": 3.1798574095292366e-06,
+      "loss": 0.0153,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6302512662627868,
+      "grad_norm": 0.3057531950463691,
+      "learning_rate": 3.176861021871267e-06,
+      "loss": 0.0109,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6304498957195352,
+      "grad_norm": 0.3058524461492556,
+      "learning_rate": 3.1738653891733228e-06,
+      "loss": 0.0102,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6306485251762837,
+      "grad_norm": 0.297644692305566,
+      "learning_rate": 3.170870512675891e-06,
+      "loss": 0.0083,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6308471546330321,
+      "grad_norm": 0.41503730094845476,
+      "learning_rate": 3.1678763936191493e-06,
+      "loss": 0.0071,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6310457840897805,
+      "grad_norm": 0.43732569514354086,
+      "learning_rate": 3.1648830332429576e-06,
+      "loss": 0.0149,
+      "step": 3177
+    },
+    {
+      "epoch": 0.6312444135465289,
+      "grad_norm": 0.4053359605257468,
+      "learning_rate": 3.161890432786864e-06,
+      "loss": 0.0119,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6314430430032774,
+      "grad_norm": 0.4018745491695557,
+      "learning_rate": 3.1588985934901024e-06,
+      "loss": 0.0102,
+      "step": 3179
+    },
+    {
+      "epoch": 0.6316416724600258,
+      "grad_norm": 0.7769830143664423,
+      "learning_rate": 3.1559075165915897e-06,
+      "loss": 0.0141,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6318403019167742,
+      "grad_norm": 0.4333501607393468,
+      "learning_rate": 3.152917203329927e-06,
+      "loss": 0.0073,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6320389313735227,
+      "grad_norm": 0.3780998689723311,
+      "learning_rate": 3.149927654943401e-06,
+      "loss": 0.0101,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6322375608302712,
+      "grad_norm": 0.5924510687322293,
+      "learning_rate": 3.14693887266998e-06,
+      "loss": 0.015,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6324361902870196,
+      "grad_norm": 0.7703336171257334,
+      "learning_rate": 3.143950857747316e-06,
+      "loss": 0.009,
+      "step": 3184
+    },
+    {
+      "epoch": 0.632634819743768,
+      "grad_norm": 0.47579749826164863,
+      "learning_rate": 3.1409636114127434e-06,
+      "loss": 0.0103,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6328334492005164,
+      "grad_norm": 0.5329585694417279,
+      "learning_rate": 3.137977134903276e-06,
+      "loss": 0.015,
+      "step": 3186
+    },
+    {
+      "epoch": 0.6330320786572649,
+      "grad_norm": 0.3926580971839605,
+      "learning_rate": 3.1349914294556146e-06,
+      "loss": 0.0122,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6332307081140133,
+      "grad_norm": 0.7331321468927112,
+      "learning_rate": 3.1320064963061335e-06,
+      "loss": 0.0179,
+      "step": 3188
+    },
+    {
+      "epoch": 0.6334293375707617,
+      "grad_norm": 0.7309606038405821,
+      "learning_rate": 3.1290223366908923e-06,
+      "loss": 0.0097,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6336279670275102,
+      "grad_norm": 0.40442647112520747,
+      "learning_rate": 3.1260389518456275e-06,
+      "loss": 0.0081,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6338265964842587,
+      "grad_norm": 0.23396485229178887,
+      "learning_rate": 3.123056343005756e-06,
+      "loss": 0.0042,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6340252259410071,
+      "grad_norm": 0.31067689461138565,
+      "learning_rate": 3.1200745114063733e-06,
+      "loss": 0.0068,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6342238553977555,
+      "grad_norm": 0.4863770217084106,
+      "learning_rate": 3.117093458282252e-06,
+      "loss": 0.0082,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6344224848545039,
+      "grad_norm": 0.6185800085213491,
+      "learning_rate": 3.1141131848678453e-06,
+      "loss": 0.0113,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6346211143112523,
+      "grad_norm": 0.4166395787409532,
+      "learning_rate": 3.111133692397279e-06,
+      "loss": 0.0072,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6348197437680008,
+      "grad_norm": 0.6441347162448549,
+      "learning_rate": 3.10815498210436e-06,
+      "loss": 0.0191,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6350183732247492,
+      "grad_norm": 0.5648143884299345,
+      "learning_rate": 3.105177055222569e-06,
+      "loss": 0.0188,
+      "step": 3197
+    },
+    {
+      "epoch": 0.6352170026814976,
+      "grad_norm": 0.2983740379639698,
+      "learning_rate": 3.102199912985061e-06,
+      "loss": 0.0074,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6354156321382461,
+      "grad_norm": 0.4602927312462115,
+      "learning_rate": 3.099223556624669e-06,
+      "loss": 0.0139,
+      "step": 3199
+    },
+    {
+      "epoch": 0.6356142615949946,
+      "grad_norm": 0.697211094443911,
+      "learning_rate": 3.096247987373897e-06,
+      "loss": 0.0155,
+      "step": 3200
+    },
+    {
+      "epoch": 0.635812891051743,
+      "grad_norm": 0.7485575048841167,
+      "learning_rate": 3.0932732064649284e-06,
+      "loss": 0.016,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6360115205084914,
+      "grad_norm": 0.78529100896383,
+      "learning_rate": 3.0902992151296156e-06,
+      "loss": 0.0226,
+      "step": 3202
+    },
+    {
+      "epoch": 0.6362101499652398,
+      "grad_norm": 0.41085026543693437,
+      "learning_rate": 3.0873260145994857e-06,
+      "loss": 0.0112,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6364087794219883,
+      "grad_norm": 0.6494964040575566,
+      "learning_rate": 3.0843536061057378e-06,
+      "loss": 0.0121,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6366074088787367,
+      "grad_norm": 0.6702929678649525,
+      "learning_rate": 3.081381990879243e-06,
+      "loss": 0.0129,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6368060383354851,
+      "grad_norm": 0.6836348848424111,
+      "learning_rate": 3.078411170150545e-06,
+      "loss": 0.0116,
+      "step": 3206
+    },
+    {
+      "epoch": 0.6370046677922336,
+      "grad_norm": 0.393253021137455,
+      "learning_rate": 3.0754411451498557e-06,
+      "loss": 0.0145,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6372032972489821,
+      "grad_norm": 1.4739247811366072,
+      "learning_rate": 3.0724719171070615e-06,
+      "loss": 0.0185,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6374019267057305,
+      "grad_norm": 0.4801500029884138,
+      "learning_rate": 3.0695034872517166e-06,
+      "loss": 0.0075,
+      "step": 3209
+    },
+    {
+      "epoch": 0.6376005561624789,
+      "grad_norm": 0.7278188114507573,
+      "learning_rate": 3.066535856813044e-06,
+      "loss": 0.0134,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6377991856192273,
+      "grad_norm": 1.2241754299310377,
+      "learning_rate": 3.063569027019936e-06,
+      "loss": 0.0163,
+      "step": 3211
+    },
+    {
+      "epoch": 0.6379978150759757,
+      "grad_norm": 0.2911942180277458,
+      "learning_rate": 3.0606029991009557e-06,
+      "loss": 0.0116,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6381964445327242,
+      "grad_norm": 0.19274138226813456,
+      "learning_rate": 3.057637774284331e-06,
+      "loss": 0.0055,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6383950739894726,
+      "grad_norm": 0.37853841414052497,
+      "learning_rate": 3.0546733537979588e-06,
+      "loss": 0.0141,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6385937034462211,
+      "grad_norm": 0.3776259956184919,
+      "learning_rate": 3.051709738869403e-06,
+      "loss": 0.0083,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6387923329029696,
+      "grad_norm": 0.560635040612453,
+      "learning_rate": 3.048746930725893e-06,
+      "loss": 0.011,
+      "step": 3216
+    },
+    {
+      "epoch": 0.638990962359718,
+      "grad_norm": 0.48130658527669296,
+      "learning_rate": 3.0457849305943256e-06,
+      "loss": 0.0118,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6391895918164664,
+      "grad_norm": 0.4771547132379289,
+      "learning_rate": 3.04282373970126e-06,
+      "loss": 0.0104,
+      "step": 3218
+    },
+    {
+      "epoch": 0.6393882212732148,
+      "grad_norm": 0.3460116166064898,
+      "learning_rate": 3.0398633592729243e-06,
+      "loss": 0.0071,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6395868507299632,
+      "grad_norm": 0.5824392990720808,
+      "learning_rate": 3.0369037905352093e-06,
+      "loss": 0.0088,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6397854801867117,
+      "grad_norm": 0.20323244555014405,
+      "learning_rate": 3.033945034713669e-06,
+      "loss": 0.0064,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6399841096434601,
+      "grad_norm": 0.638442739435459,
+      "learning_rate": 3.0309870930335204e-06,
+      "loss": 0.0137,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6401827391002085,
+      "grad_norm": 0.77301437107863,
+      "learning_rate": 3.0280299667196444e-06,
+      "loss": 0.0087,
+      "step": 3223
+    },
+    {
+      "epoch": 0.640381368556957,
+      "grad_norm": 0.7822928728867742,
+      "learning_rate": 3.0250736569965857e-06,
+      "loss": 0.0181,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6405799980137055,
+      "grad_norm": 0.33540923186909954,
+      "learning_rate": 3.0221181650885454e-06,
+      "loss": 0.0094,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6407786274704539,
+      "grad_norm": 0.3931991439665449,
+      "learning_rate": 3.0191634922193946e-06,
+      "loss": 0.014,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6409772569272023,
+      "grad_norm": 0.6751895750801706,
+      "learning_rate": 3.016209639612657e-06,
+      "loss": 0.0127,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6411758863839507,
+      "grad_norm": 0.5026416514430642,
+      "learning_rate": 3.0132566084915236e-06,
+      "loss": 0.0133,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6413745158406992,
+      "grad_norm": 0.8976233584442671,
+      "learning_rate": 3.0103044000788356e-06,
+      "loss": 0.0139,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6415731452974476,
+      "grad_norm": 0.7798871995666934,
+      "learning_rate": 3.007353015597104e-06,
+      "loss": 0.0134,
+      "step": 3230
+    },
+    {
+      "epoch": 0.641771774754196,
+      "grad_norm": 0.5380549093775145,
+      "learning_rate": 3.0044024562684938e-06,
+      "loss": 0.0139,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6419704042109445,
+      "grad_norm": 0.3154555145345817,
+      "learning_rate": 3.001452723314827e-06,
+      "loss": 0.0102,
+      "step": 3232
+    },
+    {
+      "epoch": 0.642169033667693,
+      "grad_norm": 0.3340853456189174,
+      "learning_rate": 2.998503817957587e-06,
+      "loss": 0.0058,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6423676631244414,
+      "grad_norm": 0.2936801248452502,
+      "learning_rate": 2.9955557414179117e-06,
+      "loss": 0.0043,
+      "step": 3234
+    },
+    {
+      "epoch": 0.6425662925811898,
+      "grad_norm": 0.7864359116686,
+      "learning_rate": 2.9926084949165956e-06,
+      "loss": 0.0189,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6427649220379382,
+      "grad_norm": 0.7167406061383401,
+      "learning_rate": 2.989662079674092e-06,
+      "loss": 0.0102,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6429635514946866,
+      "grad_norm": 0.3351720288278042,
+      "learning_rate": 2.9867164969105073e-06,
+      "loss": 0.0062,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6431621809514351,
+      "grad_norm": 0.40908005659181024,
+      "learning_rate": 2.983771747845606e-06,
+      "loss": 0.0111,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6433608104081835,
+      "grad_norm": 0.3277290055237293,
+      "learning_rate": 2.9808278336988043e-06,
+      "loss": 0.0073,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6435594398649319,
+      "grad_norm": 0.5048253802977973,
+      "learning_rate": 2.9778847556891754e-06,
+      "loss": 0.0077,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6437580693216804,
+      "grad_norm": 0.3405168121598885,
+      "learning_rate": 2.974942515035444e-06,
+      "loss": 0.0135,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6439566987784289,
+      "grad_norm": 0.6104144042070537,
+      "learning_rate": 2.972001112955989e-06,
+      "loss": 0.0141,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6441553282351773,
+      "grad_norm": 0.3216837585137305,
+      "learning_rate": 2.969060550668841e-06,
+      "loss": 0.0045,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6443539576919257,
+      "grad_norm": 0.4861048452267066,
+      "learning_rate": 2.966120829391686e-06,
+      "loss": 0.0073,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6445525871486741,
+      "grad_norm": 0.4499553189439919,
+      "learning_rate": 2.963181950341859e-06,
+      "loss": 0.0089,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6447512166054226,
+      "grad_norm": 0.435117209252951,
+      "learning_rate": 2.9602439147363472e-06,
+      "loss": 0.0101,
+      "step": 3246
+    },
+    {
+      "epoch": 0.644949846062171,
+      "grad_norm": 0.5927891171510099,
+      "learning_rate": 2.957306723791787e-06,
+      "loss": 0.0067,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6451484755189194,
+      "grad_norm": 0.37326850014800456,
+      "learning_rate": 2.9543703787244672e-06,
+      "loss": 0.0064,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6453471049756679,
+      "grad_norm": 0.5716478987296174,
+      "learning_rate": 2.9514348807503248e-06,
+      "loss": 0.0076,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6455457344324164,
+      "grad_norm": 0.5615102052694282,
+      "learning_rate": 2.9485002310849454e-06,
+      "loss": 0.0106,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6457443638891648,
+      "grad_norm": 0.47977193647612537,
+      "learning_rate": 2.9455664309435674e-06,
+      "loss": 0.0085,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6459429933459132,
+      "grad_norm": 0.6327506375262189,
+      "learning_rate": 2.942633481541075e-06,
+      "loss": 0.0138,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6461416228026616,
+      "grad_norm": 0.558340339517197,
+      "learning_rate": 2.9397013840919953e-06,
+      "loss": 0.0085,
+      "step": 3253
+    },
+    {
+      "epoch": 0.64634025225941,
+      "grad_norm": 0.6265421103163499,
+      "learning_rate": 2.9367701398105087e-06,
+      "loss": 0.0088,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6465388817161585,
+      "grad_norm": 0.6651718817702527,
+      "learning_rate": 2.933839749910442e-06,
+      "loss": 0.011,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6467375111729069,
+      "grad_norm": 0.7677766935582281,
+      "learning_rate": 2.930910215605265e-06,
+      "loss": 0.0126,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6469361406296554,
+      "grad_norm": 0.6946854991421672,
+      "learning_rate": 2.9279815381080966e-06,
+      "loss": 0.018,
+      "step": 3257
+    },
+    {
+      "epoch": 0.6471347700864039,
+      "grad_norm": 0.9714783676460519,
+      "learning_rate": 2.9250537186316975e-06,
+      "loss": 0.0204,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6473333995431523,
+      "grad_norm": 0.5991115101907825,
+      "learning_rate": 2.9221267583884762e-06,
+      "loss": 0.0127,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6475320289999007,
+      "grad_norm": 0.5265824803297029,
+      "learning_rate": 2.919200658590483e-06,
+      "loss": 0.0098,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6477306584566491,
+      "grad_norm": 0.4157857124333679,
+      "learning_rate": 2.9162754204494125e-06,
+      "loss": 0.0074,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6479292879133975,
+      "grad_norm": 0.5129766838144425,
+      "learning_rate": 2.913351045176606e-06,
+      "loss": 0.0135,
+      "step": 3262
+    },
+    {
+      "epoch": 0.648127917370146,
+      "grad_norm": 0.614863644307128,
+      "learning_rate": 2.91042753398304e-06,
+      "loss": 0.0082,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6483265468268944,
+      "grad_norm": 0.7205082964279731,
+      "learning_rate": 2.9075048880793395e-06,
+      "loss": 0.012,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6485251762836428,
+      "grad_norm": 0.6148645605362121,
+      "learning_rate": 2.9045831086757716e-06,
+      "loss": 0.0157,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6487238057403913,
+      "grad_norm": 1.7983560604746254,
+      "learning_rate": 2.9016621969822374e-06,
+      "loss": 0.0122,
+      "step": 3266
+    },
+    {
+      "epoch": 0.6489224351971398,
+      "grad_norm": 0.219136851702214,
+      "learning_rate": 2.8987421542082885e-06,
+      "loss": 0.0034,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6491210646538882,
+      "grad_norm": 0.6386998265055096,
+      "learning_rate": 2.8958229815631068e-06,
+      "loss": 0.0098,
+      "step": 3268
+    },
+    {
+      "epoch": 0.6493196941106366,
+      "grad_norm": 0.5134293758082206,
+      "learning_rate": 2.892904680255524e-06,
+      "loss": 0.012,
+      "step": 3269
+    },
+    {
+      "epoch": 0.649518323567385,
+      "grad_norm": 0.5345560851450427,
+      "learning_rate": 2.889987251494e-06,
+      "loss": 0.0116,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6497169530241335,
+      "grad_norm": 0.7400191306883878,
+      "learning_rate": 2.8870706964866436e-06,
+      "loss": 0.0143,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6499155824808819,
+      "grad_norm": 0.4635388451227108,
+      "learning_rate": 2.8841550164411967e-06,
+      "loss": 0.0131,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6501142119376303,
+      "grad_norm": 0.5109709585210221,
+      "learning_rate": 2.881240212565037e-06,
+      "loss": 0.0051,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6503128413943788,
+      "grad_norm": 0.7108773032847435,
+      "learning_rate": 2.878326286065185e-06,
+      "loss": 0.0139,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6505114708511273,
+      "grad_norm": 0.6626555461496284,
+      "learning_rate": 2.8754132381482926e-06,
+      "loss": 0.0119,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6507101003078757,
+      "grad_norm": 0.5269547367333896,
+      "learning_rate": 2.8725010700206514e-06,
+      "loss": 0.0145,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6509087297646241,
+      "grad_norm": 0.48445706983464565,
+      "learning_rate": 2.869589782888187e-06,
+      "loss": 0.0069,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6511073592213725,
+      "grad_norm": 0.39715086701308067,
+      "learning_rate": 2.866679377956458e-06,
+      "loss": 0.0107,
+      "step": 3278
+    },
+    {
+      "epoch": 0.651305988678121,
+      "grad_norm": 0.6713147395317994,
+      "learning_rate": 2.8637698564306637e-06,
+      "loss": 0.0145,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6515046181348694,
+      "grad_norm": 0.6460254311609365,
+      "learning_rate": 2.8608612195156318e-06,
+      "loss": 0.022,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6517032475916178,
+      "grad_norm": 0.689742162700817,
+      "learning_rate": 2.8579534684158277e-06,
+      "loss": 0.0121,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6519018770483662,
+      "grad_norm": 0.6959754949888912,
+      "learning_rate": 2.8550466043353453e-06,
+      "loss": 0.0175,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6521005065051148,
+      "grad_norm": 0.47915768218904303,
+      "learning_rate": 2.852140628477916e-06,
+      "loss": 0.0153,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6522991359618632,
+      "grad_norm": 0.6904735505096031,
+      "learning_rate": 2.849235542046904e-06,
+      "loss": 0.0146,
+      "step": 3284
+    },
+    {
+      "epoch": 0.6524977654186116,
+      "grad_norm": 0.8025885992496349,
+      "learning_rate": 2.846331346245298e-06,
+      "loss": 0.0239,
+      "step": 3285
+    },
+    {
+      "epoch": 0.65269639487536,
+      "grad_norm": 0.619059088683545,
+      "learning_rate": 2.843428042275727e-06,
+      "loss": 0.0113,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6528950243321084,
+      "grad_norm": 0.8459065891949454,
+      "learning_rate": 2.8405256313404417e-06,
+      "loss": 0.0184,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6530936537888569,
+      "grad_norm": 0.5453957141251398,
+      "learning_rate": 2.8376241146413324e-06,
+      "loss": 0.0132,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6532922832456053,
+      "grad_norm": 0.377901914882823,
+      "learning_rate": 2.8347234933799097e-06,
+      "loss": 0.0104,
+      "step": 3289
+    },
+    {
+      "epoch": 0.6534909127023537,
+      "grad_norm": 0.38612479544338135,
+      "learning_rate": 2.831823768757319e-06,
+      "loss": 0.0108,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6536895421591022,
+      "grad_norm": 0.7541376084135651,
+      "learning_rate": 2.8289249419743376e-06,
+      "loss": 0.0134,
+      "step": 3291
+    },
+    {
+      "epoch": 0.6538881716158507,
+      "grad_norm": 0.22674881778038922,
+      "learning_rate": 2.826027014231361e-06,
+      "loss": 0.0112,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6540868010725991,
+      "grad_norm": 0.5466667057789234,
+      "learning_rate": 2.8231299867284228e-06,
+      "loss": 0.0126,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6542854305293475,
+      "grad_norm": 0.5346884562381211,
+      "learning_rate": 2.820233860665175e-06,
+      "loss": 0.0085,
+      "step": 3294
+    },
+    {
+      "epoch": 0.6544840599860959,
+      "grad_norm": 0.3347131133980114,
+      "learning_rate": 2.817338637240905e-06,
+      "loss": 0.0091,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6546826894428444,
+      "grad_norm": 0.6898205612146502,
+      "learning_rate": 2.814444317654518e-06,
+      "loss": 0.01,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6548813188995928,
+      "grad_norm": 0.6113056037292642,
+      "learning_rate": 2.811550903104549e-06,
+      "loss": 0.0135,
+      "step": 3297
+    },
+    {
+      "epoch": 0.6550799483563412,
+      "grad_norm": 0.24086210100341407,
+      "learning_rate": 2.8086583947891623e-06,
+      "loss": 0.0098,
+      "step": 3298
+    },
+    {
+      "epoch": 0.6552785778130897,
+      "grad_norm": 0.29528694256953764,
+      "learning_rate": 2.8057667939061394e-06,
+      "loss": 0.0091,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6554772072698382,
+      "grad_norm": 0.32915813162257146,
+      "learning_rate": 2.8028761016528882e-06,
+      "loss": 0.0112,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6556758367265866,
+      "grad_norm": 0.5845599531100432,
+      "learning_rate": 2.7999863192264453e-06,
+      "loss": 0.0084,
+      "step": 3301
+    },
+    {
+      "epoch": 0.655874466183335,
+      "grad_norm": 0.5157143599088475,
+      "learning_rate": 2.7970974478234626e-06,
+      "loss": 0.0081,
+      "step": 3302
+    },
+    {
+      "epoch": 0.6560730956400834,
+      "grad_norm": 0.21072408406022408,
+      "learning_rate": 2.7942094886402214e-06,
+      "loss": 0.0057,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6562717250968318,
+      "grad_norm": 0.4446472530319208,
+      "learning_rate": 2.7913224428726215e-06,
+      "loss": 0.0109,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6564703545535803,
+      "grad_norm": 0.7754716523493385,
+      "learning_rate": 2.788436311716187e-06,
+      "loss": 0.0217,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6566689840103287,
+      "grad_norm": 0.3370976421618834,
+      "learning_rate": 2.78555109636606e-06,
+      "loss": 0.01,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6568676134670771,
+      "grad_norm": 0.5079691302186063,
+      "learning_rate": 2.7826667980170064e-06,
+      "loss": 0.0109,
+      "step": 3307
+    },
+    {
+      "epoch": 0.6570662429238256,
+      "grad_norm": 0.49334190860082194,
+      "learning_rate": 2.7797834178634124e-06,
+      "loss": 0.0094,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6572648723805741,
+      "grad_norm": 0.5546559533772156,
+      "learning_rate": 2.77690095709928e-06,
+      "loss": 0.0143,
+      "step": 3309
+    },
+    {
+      "epoch": 0.6574635018373225,
+      "grad_norm": 0.4708980828292352,
+      "learning_rate": 2.7740194169182377e-06,
+      "loss": 0.0122,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6576621312940709,
+      "grad_norm": 0.7391696704500784,
+      "learning_rate": 2.771138798513523e-06,
+      "loss": 0.0113,
+      "step": 3311
+    },
+    {
+      "epoch": 0.6578607607508193,
+      "grad_norm": 0.7253819426997582,
+      "learning_rate": 2.7682591030780014e-06,
+      "loss": 0.0114,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6580593902075678,
+      "grad_norm": 0.6177374318507232,
+      "learning_rate": 2.7653803318041495e-06,
+      "loss": 0.0143,
+      "step": 3313
+    },
+    {
+      "epoch": 0.6582580196643162,
+      "grad_norm": 0.43170689684000746,
+      "learning_rate": 2.7625024858840634e-06,
+      "loss": 0.0082,
+      "step": 3314
+    },
+    {
+      "epoch": 0.6584566491210646,
+      "grad_norm": 0.4260173502271494,
+      "learning_rate": 2.7596255665094594e-06,
+      "loss": 0.0125,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6586552785778131,
+      "grad_norm": 0.40502002471903215,
+      "learning_rate": 2.7567495748716632e-06,
+      "loss": 0.0127,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6588539080345616,
+      "grad_norm": 0.523810770366039,
+      "learning_rate": 2.7538745121616235e-06,
+      "loss": 0.0111,
+      "step": 3317
+    },
+    {
+      "epoch": 0.65905253749131,
+      "grad_norm": 1.0090297301535212,
+      "learning_rate": 2.751000379569897e-06,
+      "loss": 0.0136,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6592511669480584,
+      "grad_norm": 0.45840798674019173,
+      "learning_rate": 2.7481271782866623e-06,
+      "loss": 0.0045,
+      "step": 3319
+    },
+    {
+      "epoch": 0.6594497964048068,
+      "grad_norm": 0.7665200456822501,
+      "learning_rate": 2.7452549095017065e-06,
+      "loss": 0.0171,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6596484258615553,
+      "grad_norm": 0.4338808266126319,
+      "learning_rate": 2.7423835744044346e-06,
+      "loss": 0.0074,
+      "step": 3321
+    },
+    {
+      "epoch": 0.6598470553183037,
+      "grad_norm": 0.36880213167051384,
+      "learning_rate": 2.7395131741838666e-06,
+      "loss": 0.006,
+      "step": 3322
+    },
+    {
+      "epoch": 0.6600456847750521,
+      "grad_norm": 0.3898475854075473,
+      "learning_rate": 2.7366437100286286e-06,
+      "loss": 0.0088,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6602443142318005,
+      "grad_norm": 0.5107579237424404,
+      "learning_rate": 2.7337751831269637e-06,
+      "loss": 0.0132,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6604429436885491,
+      "grad_norm": 0.40112286710842526,
+      "learning_rate": 2.730907594666724e-06,
+      "loss": 0.0077,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6606415731452975,
+      "grad_norm": 0.5870623533419401,
+      "learning_rate": 2.7280409458353775e-06,
+      "loss": 0.0133,
+      "step": 3326
+    },
+    {
+      "epoch": 0.6608402026020459,
+      "grad_norm": 0.32740499713756716,
+      "learning_rate": 2.725175237820002e-06,
+      "loss": 0.0103,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6610388320587943,
+      "grad_norm": 0.4982828855799693,
+      "learning_rate": 2.72231047180728e-06,
+      "loss": 0.0075,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6612374615155427,
+      "grad_norm": 0.42218049160104415,
+      "learning_rate": 2.7194466489835132e-06,
+      "loss": 0.0128,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6614360909722912,
+      "grad_norm": 0.6155184301745691,
+      "learning_rate": 2.7165837705346033e-06,
+      "loss": 0.0155,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6616347204290396,
+      "grad_norm": 0.6852751161684462,
+      "learning_rate": 2.7137218376460683e-06,
+      "loss": 0.0093,
+      "step": 3331
+    },
+    {
+      "epoch": 0.661833349885788,
+      "grad_norm": 0.3782884028083007,
+      "learning_rate": 2.7108608515030297e-06,
+      "loss": 0.0086,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6620319793425365,
+      "grad_norm": 0.4146005267325709,
+      "learning_rate": 2.70800081329022e-06,
+      "loss": 0.0101,
+      "step": 3333
+    },
+    {
+      "epoch": 0.662230608799285,
+      "grad_norm": 0.36972782718978237,
+      "learning_rate": 2.7051417241919808e-06,
+      "loss": 0.0096,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6624292382560334,
+      "grad_norm": 0.6516852569514466,
+      "learning_rate": 2.7022835853922525e-06,
+      "loss": 0.0114,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6626278677127818,
+      "grad_norm": 0.29398406434321445,
+      "learning_rate": 2.699426398074593e-06,
+      "loss": 0.005,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6628264971695302,
+      "grad_norm": 0.3956847791060924,
+      "learning_rate": 2.6965701634221566e-06,
+      "loss": 0.0093,
+      "step": 3337
+    },
+    {
+      "epoch": 0.6630251266262787,
+      "grad_norm": 0.688077347911631,
+      "learning_rate": 2.6937148826177095e-06,
+      "loss": 0.0154,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6632237560830271,
+      "grad_norm": 0.5079820743351281,
+      "learning_rate": 2.690860556843619e-06,
+      "loss": 0.0105,
+      "step": 3339
+    },
+    {
+      "epoch": 0.6634223855397755,
+      "grad_norm": 0.4440992131204025,
+      "learning_rate": 2.688007187281859e-06,
+      "loss": 0.0102,
+      "step": 3340
+    },
+    {
+      "epoch": 0.663621014996524,
+      "grad_norm": 0.49214330396861955,
+      "learning_rate": 2.68515477511401e-06,
+      "loss": 0.0118,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6638196444532725,
+      "grad_norm": 1.0900456782512908,
+      "learning_rate": 2.6823033215212478e-06,
+      "loss": 0.022,
+      "step": 3342
+    },
+    {
+      "epoch": 0.6640182739100209,
+      "grad_norm": 0.5381851320515335,
+      "learning_rate": 2.679452827684362e-06,
+      "loss": 0.0101,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6642169033667693,
+      "grad_norm": 0.4016766272309056,
+      "learning_rate": 2.676603294783734e-06,
+      "loss": 0.0087,
+      "step": 3344
+    },
+    {
+      "epoch": 0.6644155328235177,
+      "grad_norm": 0.6652421012666038,
+      "learning_rate": 2.6737547239993565e-06,
+      "loss": 0.0186,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6646141622802662,
+      "grad_norm": 0.6608874683072442,
+      "learning_rate": 2.670907116510817e-06,
+      "loss": 0.0132,
+      "step": 3346
+    },
+    {
+      "epoch": 0.6648127917370146,
+      "grad_norm": 0.47308686328108424,
+      "learning_rate": 2.668060473497309e-06,
+      "loss": 0.0075,
+      "step": 3347
+    },
+    {
+      "epoch": 0.665011421193763,
+      "grad_norm": 0.5810128092238643,
+      "learning_rate": 2.6652147961376253e-06,
+      "loss": 0.0138,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6652100506505114,
+      "grad_norm": 0.535033001330009,
+      "learning_rate": 2.662370085610153e-06,
+      "loss": 0.0094,
+      "step": 3349
+    },
+    {
+      "epoch": 0.66540868010726,
+      "grad_norm": 0.3123641456923419,
+      "learning_rate": 2.6595263430928874e-06,
+      "loss": 0.0086,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6656073095640084,
+      "grad_norm": 0.8214580684647604,
+      "learning_rate": 2.656683569763422e-06,
+      "loss": 0.0191,
+      "step": 3351
+    },
+    {
+      "epoch": 0.6658059390207568,
+      "grad_norm": 0.8066452608717822,
+      "learning_rate": 2.6538417667989414e-06,
+      "loss": 0.0159,
+      "step": 3352
+    },
+    {
+      "epoch": 0.6660045684775052,
+      "grad_norm": 0.41875862549545895,
+      "learning_rate": 2.651000935376238e-06,
+      "loss": 0.0066,
+      "step": 3353
+    },
+    {
+      "epoch": 0.6662031979342536,
+      "grad_norm": 0.6547577856421022,
+      "learning_rate": 2.6481610766716926e-06,
+      "loss": 0.0115,
+      "step": 3354
+    },
+    {
+      "epoch": 0.6664018273910021,
+      "grad_norm": 0.6959330954668181,
+      "learning_rate": 2.6453221918612915e-06,
+      "loss": 0.014,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6666004568477505,
+      "grad_norm": 0.5030885709048069,
+      "learning_rate": 2.642484282120612e-06,
+      "loss": 0.0095,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6667990863044989,
+      "grad_norm": 0.4719568465097182,
+      "learning_rate": 2.63964734862483e-06,
+      "loss": 0.014,
+      "step": 3357
+    },
+    {
+      "epoch": 0.6669977157612474,
+      "grad_norm": 0.6796444534634053,
+      "learning_rate": 2.6368113925487182e-06,
+      "loss": 0.0091,
+      "step": 3358
+    },
+    {
+      "epoch": 0.6671963452179959,
+      "grad_norm": 0.6346635781019123,
+      "learning_rate": 2.6339764150666414e-06,
+      "loss": 0.0174,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6673949746747443,
+      "grad_norm": 0.5601756578875082,
+      "learning_rate": 2.6311424173525636e-06,
+      "loss": 0.0081,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6675936041314927,
+      "grad_norm": 0.4608080766660191,
+      "learning_rate": 2.628309400580036e-06,
+      "loss": 0.0105,
+      "step": 3361
+    },
+    {
+      "epoch": 0.6677922335882411,
+      "grad_norm": 0.8768255616150576,
+      "learning_rate": 2.6254773659222123e-06,
+      "loss": 0.0146,
+      "step": 3362
+    },
+    {
+      "epoch": 0.6679908630449896,
+      "grad_norm": 0.5125903666857362,
+      "learning_rate": 2.622646314551832e-06,
+      "loss": 0.017,
+      "step": 3363
+    },
+    {
+      "epoch": 0.668189492501738,
+      "grad_norm": 0.4250565635417071,
+      "learning_rate": 2.6198162476412324e-06,
+      "loss": 0.0168,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6683881219584864,
+      "grad_norm": 0.5993238921833851,
+      "learning_rate": 2.6169871663623424e-06,
+      "loss": 0.0079,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6685867514152348,
+      "grad_norm": 0.4691003256708194,
+      "learning_rate": 2.614159071886679e-06,
+      "loss": 0.0194,
+      "step": 3366
+    },
+    {
+      "epoch": 0.6687853808719834,
+      "grad_norm": 0.34795435504024236,
+      "learning_rate": 2.6113319653853565e-06,
+      "loss": 0.01,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6689840103287318,
+      "grad_norm": 0.7334349474214593,
+      "learning_rate": 2.6085058480290724e-06,
+      "loss": 0.0168,
+      "step": 3368
+    },
+    {
+      "epoch": 0.6691826397854802,
+      "grad_norm": 0.3782004612022668,
+      "learning_rate": 2.6056807209881247e-06,
+      "loss": 0.0097,
+      "step": 3369
+    },
+    {
+      "epoch": 0.6693812692422286,
+      "grad_norm": 0.6207564025203681,
+      "learning_rate": 2.6028565854323905e-06,
+      "loss": 0.0145,
+      "step": 3370
+    },
+    {
+      "epoch": 0.669579898698977,
+      "grad_norm": 0.3664793557426668,
+      "learning_rate": 2.6000334425313455e-06,
+      "loss": 0.0094,
+      "step": 3371
+    },
+    {
+      "epoch": 0.6697785281557255,
+      "grad_norm": 0.7947514445887053,
+      "learning_rate": 2.5972112934540483e-06,
+      "loss": 0.0137,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6699771576124739,
+      "grad_norm": 0.5244578943642776,
+      "learning_rate": 2.5943901393691463e-06,
+      "loss": 0.0146,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6701757870692223,
+      "grad_norm": 0.31429941997048494,
+      "learning_rate": 2.5915699814448813e-06,
+      "loss": 0.0084,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6703744165259709,
+      "grad_norm": 0.36243016062186306,
+      "learning_rate": 2.5887508208490726e-06,
+      "loss": 0.0089,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6705730459827193,
+      "grad_norm": 0.5468920277671807,
+      "learning_rate": 2.585932658749134e-06,
+      "loss": 0.0128,
+      "step": 3376
+    },
+    {
+      "epoch": 0.6707716754394677,
+      "grad_norm": 0.4530603705172164,
+      "learning_rate": 2.583115496312066e-06,
+      "loss": 0.0095,
+      "step": 3377
+    },
+    {
+      "epoch": 0.6709703048962161,
+      "grad_norm": 0.6766180998941852,
+      "learning_rate": 2.5802993347044494e-06,
+      "loss": 0.0111,
+      "step": 3378
+    },
+    {
+      "epoch": 0.6711689343529645,
+      "grad_norm": 0.34837311844564506,
+      "learning_rate": 2.5774841750924564e-06,
+      "loss": 0.0118,
+      "step": 3379
+    },
+    {
+      "epoch": 0.671367563809713,
+      "grad_norm": 0.4237238100703997,
+      "learning_rate": 2.5746700186418388e-06,
+      "loss": 0.0094,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6715661932664614,
+      "grad_norm": 0.2705398980966825,
+      "learning_rate": 2.5718568665179394e-06,
+      "loss": 0.0094,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6717648227232098,
+      "grad_norm": 0.7486608005177356,
+      "learning_rate": 2.5690447198856784e-06,
+      "loss": 0.0155,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6719634521799583,
+      "grad_norm": 0.38166376147207953,
+      "learning_rate": 2.566233579909564e-06,
+      "loss": 0.0082,
+      "step": 3383
+    },
+    {
+      "epoch": 0.6721620816367068,
+      "grad_norm": 0.7618562250325333,
+      "learning_rate": 2.5634234477536894e-06,
+      "loss": 0.0117,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6723607110934552,
+      "grad_norm": 0.4237062506104705,
+      "learning_rate": 2.5606143245817227e-06,
+      "loss": 0.0159,
+      "step": 3385
+    },
+    {
+      "epoch": 0.6725593405502036,
+      "grad_norm": 0.4786823768082389,
+      "learning_rate": 2.5578062115569234e-06,
+      "loss": 0.0133,
+      "step": 3386
+    },
+    {
+      "epoch": 0.672757970006952,
+      "grad_norm": 0.9433799903017728,
+      "learning_rate": 2.5549991098421245e-06,
+      "loss": 0.0195,
+      "step": 3387
+    },
+    {
+      "epoch": 0.6729565994637005,
+      "grad_norm": 0.3981282726321889,
+      "learning_rate": 2.5521930205997476e-06,
+      "loss": 0.0109,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6731552289204489,
+      "grad_norm": 0.5626283367420652,
+      "learning_rate": 2.5493879449917875e-06,
+      "loss": 0.014,
+      "step": 3389
+    },
+    {
+      "epoch": 0.6733538583771973,
+      "grad_norm": 1.4117855646706072,
+      "learning_rate": 2.546583884179825e-06,
+      "loss": 0.0172,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6735524878339457,
+      "grad_norm": 0.6383343978861289,
+      "learning_rate": 2.543780839325022e-06,
+      "loss": 0.0106,
+      "step": 3391
+    },
+    {
+      "epoch": 0.6737511172906943,
+      "grad_norm": 0.7138217958976034,
+      "learning_rate": 2.540978811588111e-06,
+      "loss": 0.0106,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6739497467474427,
+      "grad_norm": 0.34623408436167746,
+      "learning_rate": 2.5381778021294133e-06,
+      "loss": 0.0091,
+      "step": 3393
+    },
+    {
+      "epoch": 0.6741483762041911,
+      "grad_norm": 0.8557016580449002,
+      "learning_rate": 2.535377812108821e-06,
+      "loss": 0.0151,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6743470056609395,
+      "grad_norm": 0.24818536799482643,
+      "learning_rate": 2.5325788426858106e-06,
+      "loss": 0.0073,
+      "step": 3395
+    },
+    {
+      "epoch": 0.674545635117688,
+      "grad_norm": 0.31460883387664695,
+      "learning_rate": 2.5297808950194303e-06,
+      "loss": 0.0092,
+      "step": 3396
+    },
+    {
+      "epoch": 0.6747442645744364,
+      "grad_norm": 0.9971267660879034,
+      "learning_rate": 2.526983970268305e-06,
+      "loss": 0.0162,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6749428940311848,
+      "grad_norm": 0.9119079346987756,
+      "learning_rate": 2.524188069590644e-06,
+      "loss": 0.0153,
+      "step": 3398
+    },
+    {
+      "epoch": 0.6751415234879332,
+      "grad_norm": 0.7270396119139402,
+      "learning_rate": 2.521393194144222e-06,
+      "loss": 0.0131,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6753401529446817,
+      "grad_norm": 0.8376328773880279,
+      "learning_rate": 2.518599345086396e-06,
+      "loss": 0.0145,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6755387824014302,
+      "grad_norm": 0.5909829299184999,
+      "learning_rate": 2.515806523574098e-06,
+      "loss": 0.0141,
+      "step": 3401
+    },
+    {
+      "epoch": 0.6757374118581786,
+      "grad_norm": 0.7347277470642827,
+      "learning_rate": 2.513014730763829e-06,
+      "loss": 0.0136,
+      "step": 3402
+    },
+    {
+      "epoch": 0.675936041314927,
+      "grad_norm": 0.5930317321813116,
+      "learning_rate": 2.5102239678116714e-06,
+      "loss": 0.0177,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6761346707716754,
+      "grad_norm": 0.3929167896627322,
+      "learning_rate": 2.507434235873274e-06,
+      "loss": 0.0087,
+      "step": 3404
+    },
+    {
+      "epoch": 0.6763333002284239,
+      "grad_norm": 0.33640967171465747,
+      "learning_rate": 2.5046455361038646e-06,
+      "loss": 0.0066,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6765319296851723,
+      "grad_norm": 0.32477500635650375,
+      "learning_rate": 2.5018578696582387e-06,
+      "loss": 0.0108,
+      "step": 3406
+    },
+    {
+      "epoch": 0.6767305591419207,
+      "grad_norm": 0.6222872814600147,
+      "learning_rate": 2.499071237690766e-06,
+      "loss": 0.0153,
+      "step": 3407
+    },
+    {
+      "epoch": 0.6769291885986691,
+      "grad_norm": 0.3552835031167747,
+      "learning_rate": 2.496285641355392e-06,
+      "loss": 0.0054,
+      "step": 3408
+    },
+    {
+      "epoch": 0.6771278180554177,
+      "grad_norm": 0.4119535077595573,
+      "learning_rate": 2.493501081805625e-06,
+      "loss": 0.0088,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6773264475121661,
+      "grad_norm": 0.6664231184464069,
+      "learning_rate": 2.490717560194551e-06,
+      "loss": 0.0107,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6775250769689145,
+      "grad_norm": 0.6170792412387821,
+      "learning_rate": 2.4879350776748214e-06,
+      "loss": 0.013,
+      "step": 3411
+    },
+    {
+      "epoch": 0.6777237064256629,
+      "grad_norm": 0.44050710440567764,
+      "learning_rate": 2.485153635398662e-06,
+      "loss": 0.009,
+      "step": 3412
+    },
+    {
+      "epoch": 0.6779223358824114,
+      "grad_norm": 1.1119697405780344,
+      "learning_rate": 2.482373234517862e-06,
+      "loss": 0.0136,
+      "step": 3413
+    },
+    {
+      "epoch": 0.6781209653391598,
+      "grad_norm": 1.0238357497490542,
+      "learning_rate": 2.479593876183784e-06,
+      "loss": 0.0126,
+      "step": 3414
+    },
+    {
+      "epoch": 0.6783195947959082,
+      "grad_norm": 0.46374993699075306,
+      "learning_rate": 2.47681556154736e-06,
+      "loss": 0.014,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6785182242526566,
+      "grad_norm": 0.7825050148026104,
+      "learning_rate": 2.4740382917590834e-06,
+      "loss": 0.0149,
+      "step": 3416
+    },
+    {
+      "epoch": 0.6787168537094052,
+      "grad_norm": 0.45223615016890584,
+      "learning_rate": 2.4712620679690223e-06,
+      "loss": 0.0082,
+      "step": 3417
+    },
+    {
+      "epoch": 0.6789154831661536,
+      "grad_norm": 0.5359050703153959,
+      "learning_rate": 2.468486891326805e-06,
+      "loss": 0.0103,
+      "step": 3418
+    },
+    {
+      "epoch": 0.679114112622902,
+      "grad_norm": 0.6388123076656673,
+      "learning_rate": 2.4657127629816323e-06,
+      "loss": 0.0128,
+      "step": 3419
+    },
+    {
+      "epoch": 0.6793127420796504,
+      "grad_norm": 0.624704597446212,
+      "learning_rate": 2.4629396840822665e-06,
+      "loss": 0.0112,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6795113715363988,
+      "grad_norm": 0.8921444577470375,
+      "learning_rate": 2.4601676557770353e-06,
+      "loss": 0.0108,
+      "step": 3421
+    },
+    {
+      "epoch": 0.6797100009931473,
+      "grad_norm": 0.4586739844639518,
+      "learning_rate": 2.4573966792138363e-06,
+      "loss": 0.0054,
+      "step": 3422
+    },
+    {
+      "epoch": 0.6799086304498957,
+      "grad_norm": 0.3515731328172263,
+      "learning_rate": 2.4546267555401245e-06,
+      "loss": 0.0108,
+      "step": 3423
+    },
+    {
+      "epoch": 0.6801072599066441,
+      "grad_norm": 0.5037818639650333,
+      "learning_rate": 2.451857885902926e-06,
+      "loss": 0.0144,
+      "step": 3424
+    },
+    {
+      "epoch": 0.6803058893633926,
+      "grad_norm": 0.7695438647313355,
+      "learning_rate": 2.449090071448823e-06,
+      "loss": 0.0078,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6805045188201411,
+      "grad_norm": 1.0299643704038546,
+      "learning_rate": 2.4463233133239668e-06,
+      "loss": 0.0113,
+      "step": 3426
+    },
+    {
+      "epoch": 0.6807031482768895,
+      "grad_norm": 0.43828329932828497,
+      "learning_rate": 2.443557612674071e-06,
+      "loss": 0.0075,
+      "step": 3427
+    },
+    {
+      "epoch": 0.6809017777336379,
+      "grad_norm": 0.35986529838291326,
+      "learning_rate": 2.4407929706444067e-06,
+      "loss": 0.0091,
+      "step": 3428
+    },
+    {
+      "epoch": 0.6811004071903863,
+      "grad_norm": 0.5825951519865635,
+      "learning_rate": 2.4380293883798118e-06,
+      "loss": 0.0132,
+      "step": 3429
+    },
+    {
+      "epoch": 0.6812990366471348,
+      "grad_norm": 0.7054553585094656,
+      "learning_rate": 2.435266867024679e-06,
+      "loss": 0.0132,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6814976661038832,
+      "grad_norm": 0.5757263663466883,
+      "learning_rate": 2.432505407722971e-06,
+      "loss": 0.0135,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6816962955606316,
+      "grad_norm": 0.5381046623539257,
+      "learning_rate": 2.4297450116182003e-06,
+      "loss": 0.0116,
+      "step": 3432
+    },
+    {
+      "epoch": 0.68189492501738,
+      "grad_norm": 0.20663755135436665,
+      "learning_rate": 2.4269856798534463e-06,
+      "loss": 0.0044,
+      "step": 3433
+    },
+    {
+      "epoch": 0.6820935544741286,
+      "grad_norm": 0.414472597291746,
+      "learning_rate": 2.4242274135713477e-06,
+      "loss": 0.0122,
+      "step": 3434
+    },
+    {
+      "epoch": 0.682292183930877,
+      "grad_norm": 0.4289247107743778,
+      "learning_rate": 2.4214702139140954e-06,
+      "loss": 0.0103,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6824908133876254,
+      "grad_norm": 0.3110902803437075,
+      "learning_rate": 2.418714082023448e-06,
+      "loss": 0.0078,
+      "step": 3436
+    },
+    {
+      "epoch": 0.6826894428443738,
+      "grad_norm": 0.44667625859781807,
+      "learning_rate": 2.4159590190407113e-06,
+      "loss": 0.0091,
+      "step": 3437
+    },
+    {
+      "epoch": 0.6828880723011223,
+      "grad_norm": 0.6388858453233094,
+      "learning_rate": 2.4132050261067595e-06,
+      "loss": 0.0113,
+      "step": 3438
+    },
+    {
+      "epoch": 0.6830867017578707,
+      "grad_norm": 0.6488255956688261,
+      "learning_rate": 2.410452104362014e-06,
+      "loss": 0.0139,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6832853312146191,
+      "grad_norm": 0.399390098306922,
+      "learning_rate": 2.407700254946459e-06,
+      "loss": 0.0106,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6834839606713675,
+      "grad_norm": 0.42623686761820223,
+      "learning_rate": 2.404949478999634e-06,
+      "loss": 0.0114,
+      "step": 3441
+    },
+    {
+      "epoch": 0.683682590128116,
+      "grad_norm": 0.6585189499135641,
+      "learning_rate": 2.4021997776606294e-06,
+      "loss": 0.0164,
+      "step": 3442
+    },
+    {
+      "epoch": 0.6838812195848645,
+      "grad_norm": 0.5067746225527527,
+      "learning_rate": 2.3994511520680975e-06,
+      "loss": 0.0229,
+      "step": 3443
+    },
+    {
+      "epoch": 0.6840798490416129,
+      "grad_norm": 0.6378096505999058,
+      "learning_rate": 2.3967036033602388e-06,
+      "loss": 0.0119,
+      "step": 3444
+    },
+    {
+      "epoch": 0.6842784784983613,
+      "grad_norm": 0.5093697789433972,
+      "learning_rate": 2.393957132674809e-06,
+      "loss": 0.0111,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6844771079551097,
+      "grad_norm": 0.8605508533457091,
+      "learning_rate": 2.3912117411491233e-06,
+      "loss": 0.0123,
+      "step": 3446
+    },
+    {
+      "epoch": 0.6846757374118582,
+      "grad_norm": 0.3992536191232021,
+      "learning_rate": 2.3884674299200404e-06,
+      "loss": 0.0073,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6848743668686066,
+      "grad_norm": 0.3301132162498039,
+      "learning_rate": 2.3857242001239815e-06,
+      "loss": 0.0111,
+      "step": 3448
+    },
+    {
+      "epoch": 0.685072996325355,
+      "grad_norm": 0.47428147751600336,
+      "learning_rate": 2.3829820528969118e-06,
+      "loss": 0.0149,
+      "step": 3449
+    },
+    {
+      "epoch": 0.6852716257821034,
+      "grad_norm": 0.5910659829970917,
+      "learning_rate": 2.3802409893743534e-06,
+      "loss": 0.0128,
+      "step": 3450
+    },
+    {
+      "epoch": 0.685470255238852,
+      "grad_norm": 0.3809691879291683,
+      "learning_rate": 2.3775010106913794e-06,
+      "loss": 0.0128,
+      "step": 3451
+    },
+    {
+      "epoch": 0.6856688846956004,
+      "grad_norm": 0.5463503817648073,
+      "learning_rate": 2.3747621179826084e-06,
+      "loss": 0.0109,
+      "step": 3452
+    },
+    {
+      "epoch": 0.6858675141523488,
+      "grad_norm": 0.5779042575602321,
+      "learning_rate": 2.3720243123822177e-06,
+      "loss": 0.0154,
+      "step": 3453
+    },
+    {
+      "epoch": 0.6860661436090972,
+      "grad_norm": 0.5292906640923668,
+      "learning_rate": 2.3692875950239256e-06,
+      "loss": 0.012,
+      "step": 3454
+    },
+    {
+      "epoch": 0.6862647730658457,
+      "grad_norm": 0.7193268858083541,
+      "learning_rate": 2.366551967041008e-06,
+      "loss": 0.015,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6864634025225941,
+      "grad_norm": 0.7981922366337875,
+      "learning_rate": 2.3638174295662815e-06,
+      "loss": 0.0149,
+      "step": 3456
+    },
+    {
+      "epoch": 0.6866620319793425,
+      "grad_norm": 1.8969690617361947,
+      "learning_rate": 2.3610839837321174e-06,
+      "loss": 0.009,
+      "step": 3457
+    },
+    {
+      "epoch": 0.6868606614360909,
+      "grad_norm": 1.1223983859989493,
+      "learning_rate": 2.3583516306704353e-06,
+      "loss": 0.0138,
+      "step": 3458
+    },
+    {
+      "epoch": 0.6870592908928395,
+      "grad_norm": 0.7263132364715444,
+      "learning_rate": 2.3556203715126953e-06,
+      "loss": 0.0209,
+      "step": 3459
+    },
+    {
+      "epoch": 0.6872579203495879,
+      "grad_norm": 0.4466445789942108,
+      "learning_rate": 2.352890207389913e-06,
+      "loss": 0.0097,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6874565498063363,
+      "grad_norm": 0.9150117897195746,
+      "learning_rate": 2.3501611394326434e-06,
+      "loss": 0.0133,
+      "step": 3461
+    },
+    {
+      "epoch": 0.6876551792630847,
+      "grad_norm": 0.5168870032711956,
+      "learning_rate": 2.3474331687709937e-06,
+      "loss": 0.0136,
+      "step": 3462
+    },
+    {
+      "epoch": 0.6878538087198331,
+      "grad_norm": 0.31968699277456936,
+      "learning_rate": 2.3447062965346113e-06,
+      "loss": 0.0073,
+      "step": 3463
+    },
+    {
+      "epoch": 0.6880524381765816,
+      "grad_norm": 0.6109378846286353,
+      "learning_rate": 2.3419805238526923e-06,
+      "loss": 0.0179,
+      "step": 3464
+    },
+    {
+      "epoch": 0.68825106763333,
+      "grad_norm": 0.35760154395629234,
+      "learning_rate": 2.339255851853978e-06,
+      "loss": 0.0115,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6884496970900784,
+      "grad_norm": 0.5708486697315434,
+      "learning_rate": 2.3365322816667497e-06,
+      "loss": 0.0132,
+      "step": 3466
+    },
+    {
+      "epoch": 0.688648326546827,
+      "grad_norm": 0.6178889637099267,
+      "learning_rate": 2.3338098144188394e-06,
+      "loss": 0.0093,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6888469560035754,
+      "grad_norm": 0.16097938067312248,
+      "learning_rate": 2.3310884512376115e-06,
+      "loss": 0.0034,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6890455854603238,
+      "grad_norm": 0.3961429560117535,
+      "learning_rate": 2.328368193249983e-06,
+      "loss": 0.0067,
+      "step": 3469
+    },
+    {
+      "epoch": 0.6892442149170722,
+      "grad_norm": 0.38510183274734977,
+      "learning_rate": 2.3256490415824124e-06,
+      "loss": 0.0118,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6894428443738206,
+      "grad_norm": 0.9257497192314168,
+      "learning_rate": 2.322930997360894e-06,
+      "loss": 0.014,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6896414738305691,
+      "grad_norm": 0.6459018691049223,
+      "learning_rate": 2.3202140617109697e-06,
+      "loss": 0.0098,
+      "step": 3472
+    },
+    {
+      "epoch": 0.6898401032873175,
+      "grad_norm": 1.043612774948338,
+      "learning_rate": 2.3174982357577174e-06,
+      "loss": 0.0099,
+      "step": 3473
+    },
+    {
+      "epoch": 0.6900387327440659,
+      "grad_norm": 0.2929171606322306,
+      "learning_rate": 2.3147835206257615e-06,
+      "loss": 0.0066,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6902373622008143,
+      "grad_norm": 0.4205660357282004,
+      "learning_rate": 2.312069917439258e-06,
+      "loss": 0.0106,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6904359916575629,
+      "grad_norm": 0.28539942824667136,
+      "learning_rate": 2.3093574273219103e-06,
+      "loss": 0.0107,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6906346211143113,
+      "grad_norm": 0.38833687964976527,
+      "learning_rate": 2.30664605139696e-06,
+      "loss": 0.0107,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6908332505710597,
+      "grad_norm": 0.40899264773370975,
+      "learning_rate": 2.30393579078718e-06,
+      "loss": 0.0044,
+      "step": 3478
+    },
+    {
+      "epoch": 0.6910318800278081,
+      "grad_norm": 0.5323829934882054,
+      "learning_rate": 2.3012266466148913e-06,
+      "loss": 0.0164,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6912305094845566,
+      "grad_norm": 0.5241906848713265,
+      "learning_rate": 2.2985186200019434e-06,
+      "loss": 0.0091,
+      "step": 3480
+    },
+    {
+      "epoch": 0.691429138941305,
+      "grad_norm": 0.47844187118354103,
+      "learning_rate": 2.2958117120697322e-06,
+      "loss": 0.0088,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6916277683980534,
+      "grad_norm": 0.5162909926785495,
+      "learning_rate": 2.29310592393918e-06,
+      "loss": 0.0079,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6918263978548018,
+      "grad_norm": 0.4668928613690837,
+      "learning_rate": 2.290401256730755e-06,
+      "loss": 0.0087,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6920250273115504,
+      "grad_norm": 0.47892066762124963,
+      "learning_rate": 2.2876977115644577e-06,
+      "loss": 0.0098,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6922236567682988,
+      "grad_norm": 0.4015035321499451,
+      "learning_rate": 2.28499528955982e-06,
+      "loss": 0.0082,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6924222862250472,
+      "grad_norm": 0.4774526229757127,
+      "learning_rate": 2.2822939918359157e-06,
+      "loss": 0.0131,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6926209156817956,
+      "grad_norm": 0.3237800455442647,
+      "learning_rate": 2.279593819511346e-06,
+      "loss": 0.0094,
+      "step": 3487
+    },
+    {
+      "epoch": 0.692819545138544,
+      "grad_norm": 0.8338717940603173,
+      "learning_rate": 2.276894773704253e-06,
+      "loss": 0.0164,
+      "step": 3488
+    },
+    {
+      "epoch": 0.6930181745952925,
+      "grad_norm": 0.3897033888651404,
+      "learning_rate": 2.274196855532306e-06,
+      "loss": 0.0061,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6932168040520409,
+      "grad_norm": 0.6207140161695185,
+      "learning_rate": 2.2715000661127107e-06,
+      "loss": 0.0168,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6934154335087893,
+      "grad_norm": 0.36947534908833934,
+      "learning_rate": 2.26880440656221e-06,
+      "loss": 0.0092,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6936140629655377,
+      "grad_norm": 0.5230707821198057,
+      "learning_rate": 2.2661098779970653e-06,
+      "loss": 0.0116,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6938126924222863,
+      "grad_norm": 0.34175221266600403,
+      "learning_rate": 2.263416481533082e-06,
+      "loss": 0.0057,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6940113218790347,
+      "grad_norm": 0.5044214335668877,
+      "learning_rate": 2.260724218285596e-06,
+      "loss": 0.0121,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6942099513357831,
+      "grad_norm": 0.7343017142175041,
+      "learning_rate": 2.258033089369466e-06,
+      "loss": 0.0109,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6944085807925315,
+      "grad_norm": 0.7027496441081319,
+      "learning_rate": 2.255343095899089e-06,
+      "loss": 0.0133,
+      "step": 3496
+    },
+    {
+      "epoch": 0.69460721024928,
+      "grad_norm": 0.515249317298665,
+      "learning_rate": 2.252654238988386e-06,
+      "loss": 0.0117,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6948058397060284,
+      "grad_norm": 0.3406590742309431,
+      "learning_rate": 2.2499665197508137e-06,
+      "loss": 0.0073,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6950044691627768,
+      "grad_norm": 1.054212903506879,
+      "learning_rate": 2.247279939299349e-06,
+      "loss": 0.0198,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6952030986195252,
+      "grad_norm": 1.254784991849465,
+      "learning_rate": 2.244594498746505e-06,
+      "loss": 0.0131,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6954017280762738,
+      "grad_norm": 0.5023270239108601,
+      "learning_rate": 2.241910199204321e-06,
+      "loss": 0.0068,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6956003575330222,
+      "grad_norm": 0.33395472573545676,
+      "learning_rate": 2.239227041784361e-06,
+      "loss": 0.0054,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6957989869897706,
+      "grad_norm": 1.1377173779793461,
+      "learning_rate": 2.236545027597719e-06,
+      "loss": 0.0275,
+      "step": 3503
+    },
+    {
+      "epoch": 0.695997616446519,
+      "grad_norm": 0.46319236840052935,
+      "learning_rate": 2.2338641577550124e-06,
+      "loss": 0.0162,
+      "step": 3504
+    },
+    {
+      "epoch": 0.6961962459032675,
+      "grad_norm": 0.3934534179163216,
+      "learning_rate": 2.2311844333663897e-06,
+      "loss": 0.0071,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6963948753600159,
+      "grad_norm": 0.7869903742607097,
+      "learning_rate": 2.2285058555415184e-06,
+      "loss": 0.0174,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6965935048167643,
+      "grad_norm": 0.5402812555281423,
+      "learning_rate": 2.2258284253895975e-06,
+      "loss": 0.0067,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6967921342735127,
+      "grad_norm": 0.5141673771075058,
+      "learning_rate": 2.2231521440193486e-06,
+      "loss": 0.0141,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6969907637302611,
+      "grad_norm": 0.7019261554336447,
+      "learning_rate": 2.2204770125390147e-06,
+      "loss": 0.0112,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6971893931870097,
+      "grad_norm": 0.6811176170277796,
+      "learning_rate": 2.2178030320563677e-06,
+      "loss": 0.0106,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6973880226437581,
+      "grad_norm": 0.39398429666162565,
+      "learning_rate": 2.2151302036786976e-06,
+      "loss": 0.0087,
+      "step": 3511
+    },
+    {
+      "epoch": 0.6975866521005065,
+      "grad_norm": 0.7331468400657002,
+      "learning_rate": 2.212458528512823e-06,
+      "loss": 0.0166,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6977852815572549,
+      "grad_norm": 0.5535187539164983,
+      "learning_rate": 2.2097880076650786e-06,
+      "loss": 0.0096,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6979839110140034,
+      "grad_norm": 0.3967575431172973,
+      "learning_rate": 2.2071186422413253e-06,
+      "loss": 0.0116,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6981825404707518,
+      "grad_norm": 0.7538890683797367,
+      "learning_rate": 2.2044504333469497e-06,
+      "loss": 0.0134,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6983811699275002,
+      "grad_norm": 0.3956918733926533,
+      "learning_rate": 2.2017833820868466e-06,
+      "loss": 0.0119,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6985797993842486,
+      "grad_norm": 0.6929067669197901,
+      "learning_rate": 2.1991174895654445e-06,
+      "loss": 0.0131,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6987784288409972,
+      "grad_norm": 3.3273773145583596,
+      "learning_rate": 2.1964527568866828e-06,
+      "loss": 0.013,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6989770582977456,
+      "grad_norm": 0.9093484745445498,
+      "learning_rate": 2.193789185154027e-06,
+      "loss": 0.0223,
+      "step": 3519
+    },
+    {
+      "epoch": 0.699175687754494,
+      "grad_norm": 0.6123078414872011,
+      "learning_rate": 2.191126775470462e-06,
+      "loss": 0.0098,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6993743172112424,
+      "grad_norm": 0.5104306223926839,
+      "learning_rate": 2.188465528938484e-06,
+      "loss": 0.0169,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6995729466679909,
+      "grad_norm": 0.6464141115145141,
+      "learning_rate": 2.185805446660117e-06,
+      "loss": 0.0158,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6997715761247393,
+      "grad_norm": 0.5539549495565598,
+      "learning_rate": 2.183146529736894e-06,
+      "loss": 0.0142,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6999702055814877,
+      "grad_norm": 1.1776865246831185,
+      "learning_rate": 2.180488779269874e-06,
+      "loss": 0.0158,
+      "step": 3524
+    },
+    {
+      "epoch": 0.7001688350382361,
+      "grad_norm": 0.5566688947877497,
+      "learning_rate": 2.177832196359626e-06,
+      "loss": 0.0093,
+      "step": 3525
+    },
+    {
+      "epoch": 0.7003674644949847,
+      "grad_norm": 0.2616688972825115,
+      "learning_rate": 2.1751767821062385e-06,
+      "loss": 0.0077,
+      "step": 3526
+    },
+    {
+      "epoch": 0.7005660939517331,
+      "grad_norm": 0.8922237616752913,
+      "learning_rate": 2.172522537609319e-06,
+      "loss": 0.0136,
+      "step": 3527
+    },
+    {
+      "epoch": 0.7007647234084815,
+      "grad_norm": 0.42603522315846554,
+      "learning_rate": 2.169869463967983e-06,
+      "loss": 0.0178,
+      "step": 3528
+    },
+    {
+      "epoch": 0.7009633528652299,
+      "grad_norm": 0.4466000675946425,
+      "learning_rate": 2.16721756228087e-06,
+      "loss": 0.0137,
+      "step": 3529
+    },
+    {
+      "epoch": 0.7011619823219783,
+      "grad_norm": 0.6276370519416008,
+      "learning_rate": 2.164566833646125e-06,
+      "loss": 0.0192,
+      "step": 3530
+    },
+    {
+      "epoch": 0.7013606117787268,
+      "grad_norm": 0.5258723027999391,
+      "learning_rate": 2.1619172791614175e-06,
+      "loss": 0.0151,
+      "step": 3531
+    },
+    {
+      "epoch": 0.7015592412354752,
+      "grad_norm": 0.6959177505640589,
+      "learning_rate": 2.1592688999239204e-06,
+      "loss": 0.0166,
+      "step": 3532
+    },
+    {
+      "epoch": 0.7017578706922236,
+      "grad_norm": 0.6402652235456184,
+      "learning_rate": 2.156621697030327e-06,
+      "loss": 0.0114,
+      "step": 3533
+    },
+    {
+      "epoch": 0.701956500148972,
+      "grad_norm": 0.46310938800434814,
+      "learning_rate": 2.1539756715768434e-06,
+      "loss": 0.0097,
+      "step": 3534
+    },
+    {
+      "epoch": 0.7021551296057206,
+      "grad_norm": 0.44032965187225664,
+      "learning_rate": 2.151330824659182e-06,
+      "loss": 0.0097,
+      "step": 3535
+    },
+    {
+      "epoch": 0.702353759062469,
+      "grad_norm": 0.8920415210582037,
+      "learning_rate": 2.1486871573725752e-06,
+      "loss": 0.0143,
+      "step": 3536
+    },
+    {
+      "epoch": 0.7025523885192174,
+      "grad_norm": 0.37431884202742494,
+      "learning_rate": 2.1460446708117594e-06,
+      "loss": 0.0098,
+      "step": 3537
+    },
+    {
+      "epoch": 0.7027510179759658,
+      "grad_norm": 0.4891117814991983,
+      "learning_rate": 2.1434033660709896e-06,
+      "loss": 0.0131,
+      "step": 3538
+    },
+    {
+      "epoch": 0.7029496474327143,
+      "grad_norm": 1.047496707565232,
+      "learning_rate": 2.1407632442440247e-06,
+      "loss": 0.0132,
+      "step": 3539
+    },
+    {
+      "epoch": 0.7031482768894627,
+      "grad_norm": 0.4074516687437718,
+      "learning_rate": 2.138124306424136e-06,
+      "loss": 0.0096,
+      "step": 3540
+    },
+    {
+      "epoch": 0.7033469063462111,
+      "grad_norm": 0.6921084878895216,
+      "learning_rate": 2.135486553704107e-06,
+      "loss": 0.012,
+      "step": 3541
+    },
+    {
+      "epoch": 0.7035455358029595,
+      "grad_norm": 0.352632241603671,
+      "learning_rate": 2.132849987176226e-06,
+      "loss": 0.0085,
+      "step": 3542
+    },
+    {
+      "epoch": 0.7037441652597081,
+      "grad_norm": 0.5517973185214629,
+      "learning_rate": 2.1302146079322945e-06,
+      "loss": 0.0078,
+      "step": 3543
+    },
+    {
+      "epoch": 0.7039427947164565,
+      "grad_norm": 0.3723571720752435,
+      "learning_rate": 2.1275804170636213e-06,
+      "loss": 0.0087,
+      "step": 3544
+    },
+    {
+      "epoch": 0.7041414241732049,
+      "grad_norm": 0.57601818247446,
+      "learning_rate": 2.124947415661019e-06,
+      "loss": 0.0152,
+      "step": 3545
+    },
+    {
+      "epoch": 0.7043400536299533,
+      "grad_norm": 0.21023435616038813,
+      "learning_rate": 2.1223156048148146e-06,
+      "loss": 0.0068,
+      "step": 3546
+    },
+    {
+      "epoch": 0.7045386830867018,
+      "grad_norm": 0.33170514161187187,
+      "learning_rate": 2.119684985614835e-06,
+      "loss": 0.0091,
+      "step": 3547
+    },
+    {
+      "epoch": 0.7047373125434502,
+      "grad_norm": 0.6974394026163584,
+      "learning_rate": 2.1170555591504198e-06,
+      "loss": 0.0195,
+      "step": 3548
+    },
+    {
+      "epoch": 0.7049359420001986,
+      "grad_norm": 0.6571182045495615,
+      "learning_rate": 2.1144273265104088e-06,
+      "loss": 0.0166,
+      "step": 3549
+    },
+    {
+      "epoch": 0.705134571456947,
+      "grad_norm": 0.5118877638635031,
+      "learning_rate": 2.1118002887831523e-06,
+      "loss": 0.017,
+      "step": 3550
+    },
+    {
+      "epoch": 0.7053332009136954,
+      "grad_norm": 0.607122918396555,
+      "learning_rate": 2.1091744470565062e-06,
+      "loss": 0.0123,
+      "step": 3551
+    },
+    {
+      "epoch": 0.705531830370444,
+      "grad_norm": 0.46800730184021394,
+      "learning_rate": 2.1065498024178237e-06,
+      "loss": 0.0117,
+      "step": 3552
+    },
+    {
+      "epoch": 0.7057304598271924,
+      "grad_norm": 0.48441945139013526,
+      "learning_rate": 2.1039263559539737e-06,
+      "loss": 0.0137,
+      "step": 3553
+    },
+    {
+      "epoch": 0.7059290892839408,
+      "grad_norm": 0.6602368453331714,
+      "learning_rate": 2.1013041087513163e-06,
+      "loss": 0.0202,
+      "step": 3554
+    },
+    {
+      "epoch": 0.7061277187406892,
+      "grad_norm": 0.4140800156964033,
+      "learning_rate": 2.098683061895727e-06,
+      "loss": 0.013,
+      "step": 3555
+    },
+    {
+      "epoch": 0.7063263481974377,
+      "grad_norm": 0.39987172496263784,
+      "learning_rate": 2.0960632164725746e-06,
+      "loss": 0.009,
+      "step": 3556
+    },
+    {
+      "epoch": 0.7065249776541861,
+      "grad_norm": 0.42338870067574086,
+      "learning_rate": 2.0934445735667366e-06,
+      "loss": 0.012,
+      "step": 3557
+    },
+    {
+      "epoch": 0.7067236071109345,
+      "grad_norm": 0.5296392321338635,
+      "learning_rate": 2.0908271342625907e-06,
+      "loss": 0.0116,
+      "step": 3558
+    },
+    {
+      "epoch": 0.7069222365676829,
+      "grad_norm": 0.6562700338635507,
+      "learning_rate": 2.0882108996440144e-06,
+      "loss": 0.0162,
+      "step": 3559
+    },
+    {
+      "epoch": 0.7071208660244315,
+      "grad_norm": 0.45596472145739425,
+      "learning_rate": 2.0855958707943903e-06,
+      "loss": 0.0132,
+      "step": 3560
+    },
+    {
+      "epoch": 0.7073194954811799,
+      "grad_norm": 0.5851977975386901,
+      "learning_rate": 2.082982048796595e-06,
+      "loss": 0.0115,
+      "step": 3561
+    },
+    {
+      "epoch": 0.7075181249379283,
+      "grad_norm": 0.6246161154127471,
+      "learning_rate": 2.0803694347330146e-06,
+      "loss": 0.025,
+      "step": 3562
+    },
+    {
+      "epoch": 0.7077167543946767,
+      "grad_norm": 0.70364745633366,
+      "learning_rate": 2.077758029685527e-06,
+      "loss": 0.0059,
+      "step": 3563
+    },
+    {
+      "epoch": 0.7079153838514252,
+      "grad_norm": 1.2354575585632148,
+      "learning_rate": 2.0751478347355112e-06,
+      "loss": 0.0226,
+      "step": 3564
+    },
+    {
+      "epoch": 0.7081140133081736,
+      "grad_norm": 0.40712325234995933,
+      "learning_rate": 2.0725388509638504e-06,
+      "loss": 0.0098,
+      "step": 3565
+    },
+    {
+      "epoch": 0.708312642764922,
+      "grad_norm": 0.318136163341281,
+      "learning_rate": 2.0699310794509176e-06,
+      "loss": 0.0089,
+      "step": 3566
+    },
+    {
+      "epoch": 0.7085112722216704,
+      "grad_norm": 0.22133628429597416,
+      "learning_rate": 2.0673245212765923e-06,
+      "loss": 0.0054,
+      "step": 3567
+    },
+    {
+      "epoch": 0.708709901678419,
+      "grad_norm": 1.0519287362857404,
+      "learning_rate": 2.064719177520244e-06,
+      "loss": 0.0146,
+      "step": 3568
+    },
+    {
+      "epoch": 0.7089085311351674,
+      "grad_norm": 0.38952311086071517,
+      "learning_rate": 2.062115049260745e-06,
+      "loss": 0.0088,
+      "step": 3569
+    },
+    {
+      "epoch": 0.7091071605919158,
+      "grad_norm": 0.5339933404088719,
+      "learning_rate": 2.0595121375764638e-06,
+      "loss": 0.0148,
+      "step": 3570
+    },
+    {
+      "epoch": 0.7093057900486642,
+      "grad_norm": 0.35237702102373786,
+      "learning_rate": 2.0569104435452597e-06,
+      "loss": 0.0071,
+      "step": 3571
+    },
+    {
+      "epoch": 0.7095044195054127,
+      "grad_norm": 0.5657247740451654,
+      "learning_rate": 2.0543099682444958e-06,
+      "loss": 0.0107,
+      "step": 3572
+    },
+    {
+      "epoch": 0.7097030489621611,
+      "grad_norm": 0.5548852852278983,
+      "learning_rate": 2.0517107127510223e-06,
+      "loss": 0.0107,
+      "step": 3573
+    },
+    {
+      "epoch": 0.7099016784189095,
+      "grad_norm": 0.34677930544310487,
+      "learning_rate": 2.049112678141192e-06,
+      "loss": 0.0053,
+      "step": 3574
+    },
+    {
+      "epoch": 0.7101003078756579,
+      "grad_norm": 0.3763747072298554,
+      "learning_rate": 2.0465158654908447e-06,
+      "loss": 0.0092,
+      "step": 3575
+    },
+    {
+      "epoch": 0.7102989373324063,
+      "grad_norm": 0.4447085423083739,
+      "learning_rate": 2.0439202758753196e-06,
+      "loss": 0.0177,
+      "step": 3576
+    },
+    {
+      "epoch": 0.7104975667891549,
+      "grad_norm": 0.43408562131660483,
+      "learning_rate": 2.04132591036945e-06,
+      "loss": 0.011,
+      "step": 3577
+    },
+    {
+      "epoch": 0.7106961962459033,
+      "grad_norm": 0.7457101724605041,
+      "learning_rate": 2.0387327700475564e-06,
+      "loss": 0.0103,
+      "step": 3578
+    },
+    {
+      "epoch": 0.7108948257026517,
+      "grad_norm": 0.7667181305870763,
+      "learning_rate": 2.036140855983458e-06,
+      "loss": 0.0115,
+      "step": 3579
+    },
+    {
+      "epoch": 0.7110934551594001,
+      "grad_norm": 0.4608200956966256,
+      "learning_rate": 2.0335501692504624e-06,
+      "loss": 0.0138,
+      "step": 3580
+    },
+    {
+      "epoch": 0.7112920846161486,
+      "grad_norm": 0.43968224621643054,
+      "learning_rate": 2.0309607109213725e-06,
+      "loss": 0.0119,
+      "step": 3581
+    },
+    {
+      "epoch": 0.711490714072897,
+      "grad_norm": 0.38784857298098396,
+      "learning_rate": 2.028372482068477e-06,
+      "loss": 0.0086,
+      "step": 3582
+    },
+    {
+      "epoch": 0.7116893435296454,
+      "grad_norm": 0.4987232153713858,
+      "learning_rate": 2.025785483763561e-06,
+      "loss": 0.0195,
+      "step": 3583
+    },
+    {
+      "epoch": 0.7118879729863938,
+      "grad_norm": 0.7964544674594572,
+      "learning_rate": 2.0231997170778996e-06,
+      "loss": 0.0187,
+      "step": 3584
+    },
+    {
+      "epoch": 0.7120866024431424,
+      "grad_norm": 0.9392661611065467,
+      "learning_rate": 2.0206151830822523e-06,
+      "loss": 0.013,
+      "step": 3585
+    },
+    {
+      "epoch": 0.7122852318998908,
+      "grad_norm": 0.5308227215692699,
+      "learning_rate": 2.018031882846876e-06,
+      "loss": 0.0074,
+      "step": 3586
+    },
+    {
+      "epoch": 0.7124838613566392,
+      "grad_norm": 0.44551851990638935,
+      "learning_rate": 2.0154498174415104e-06,
+      "loss": 0.0069,
+      "step": 3587
+    },
+    {
+      "epoch": 0.7126824908133876,
+      "grad_norm": 0.4912170131917664,
+      "learning_rate": 2.0128689879353853e-06,
+      "loss": 0.0249,
+      "step": 3588
+    },
+    {
+      "epoch": 0.7128811202701361,
+      "grad_norm": 0.9819312671738274,
+      "learning_rate": 2.0102893953972223e-06,
+      "loss": 0.0155,
+      "step": 3589
+    },
+    {
+      "epoch": 0.7130797497268845,
+      "grad_norm": 0.6813312541973012,
+      "learning_rate": 2.0077110408952254e-06,
+      "loss": 0.0128,
+      "step": 3590
+    },
+    {
+      "epoch": 0.7132783791836329,
+      "grad_norm": 0.5331664640007728,
+      "learning_rate": 2.0051339254970912e-06,
+      "loss": 0.0146,
+      "step": 3591
+    },
+    {
+      "epoch": 0.7134770086403813,
+      "grad_norm": 0.2943537320564372,
+      "learning_rate": 2.0025580502699983e-06,
+      "loss": 0.0083,
+      "step": 3592
+    },
+    {
+      "epoch": 0.7136756380971297,
+      "grad_norm": 0.5213864974119058,
+      "learning_rate": 1.9999834162806143e-06,
+      "loss": 0.0119,
+      "step": 3593
+    },
+    {
+      "epoch": 0.7138742675538783,
+      "grad_norm": 0.46666853895153365,
+      "learning_rate": 1.997410024595095e-06,
+      "loss": 0.0091,
+      "step": 3594
+    },
+    {
+      "epoch": 0.7140728970106267,
+      "grad_norm": 0.5085939748759494,
+      "learning_rate": 1.9948378762790767e-06,
+      "loss": 0.0138,
+      "step": 3595
+    },
+    {
+      "epoch": 0.7142715264673751,
+      "grad_norm": 0.2990427145994346,
+      "learning_rate": 1.992266972397685e-06,
+      "loss": 0.0062,
+      "step": 3596
+    },
+    {
+      "epoch": 0.7144701559241236,
+      "grad_norm": 0.7704563352878734,
+      "learning_rate": 1.9896973140155274e-06,
+      "loss": 0.0136,
+      "step": 3597
+    },
+    {
+      "epoch": 0.714668785380872,
+      "grad_norm": 0.4896227179477158,
+      "learning_rate": 1.9871289021966984e-06,
+      "loss": 0.0091,
+      "step": 3598
+    },
+    {
+      "epoch": 0.7148674148376204,
+      "grad_norm": 0.5270626574804558,
+      "learning_rate": 1.9845617380047725e-06,
+      "loss": 0.0119,
+      "step": 3599
+    },
+    {
+      "epoch": 0.7150660442943688,
+      "grad_norm": 0.38674602009064785,
+      "learning_rate": 1.981995822502811e-06,
+      "loss": 0.008,
+      "step": 3600
+    },
+    {
+      "epoch": 0.7152646737511172,
+      "grad_norm": 0.291567327736593,
+      "learning_rate": 1.979431156753359e-06,
+      "loss": 0.0089,
+      "step": 3601
+    },
+    {
+      "epoch": 0.7154633032078658,
+      "grad_norm": 0.7286287787040973,
+      "learning_rate": 1.9768677418184382e-06,
+      "loss": 0.0097,
+      "step": 3602
+    },
+    {
+      "epoch": 0.7156619326646142,
+      "grad_norm": 0.3147595174206081,
+      "learning_rate": 1.97430557875956e-06,
+      "loss": 0.0074,
+      "step": 3603
+    },
+    {
+      "epoch": 0.7158605621213626,
+      "grad_norm": 0.557758074899719,
+      "learning_rate": 1.97174466863771e-06,
+      "loss": 0.014,
+      "step": 3604
+    },
+    {
+      "epoch": 0.716059191578111,
+      "grad_norm": 0.312647733026765,
+      "learning_rate": 1.9691850125133617e-06,
+      "loss": 0.0106,
+      "step": 3605
+    },
+    {
+      "epoch": 0.7162578210348595,
+      "grad_norm": 0.5740235026460454,
+      "learning_rate": 1.9666266114464626e-06,
+      "loss": 0.0105,
+      "step": 3606
+    },
+    {
+      "epoch": 0.7164564504916079,
+      "grad_norm": 0.42789475100988844,
+      "learning_rate": 1.964069466496446e-06,
+      "loss": 0.008,
+      "step": 3607
+    },
+    {
+      "epoch": 0.7166550799483563,
+      "grad_norm": 0.3997543712505119,
+      "learning_rate": 1.961513578722225e-06,
+      "loss": 0.0067,
+      "step": 3608
+    },
+    {
+      "epoch": 0.7168537094051047,
+      "grad_norm": 0.5950672430416363,
+      "learning_rate": 1.958958949182187e-06,
+      "loss": 0.012,
+      "step": 3609
+    },
+    {
+      "epoch": 0.7170523388618533,
+      "grad_norm": 0.6623144173230123,
+      "learning_rate": 1.956405578934204e-06,
+      "loss": 0.0094,
+      "step": 3610
+    },
+    {
+      "epoch": 0.7172509683186017,
+      "grad_norm": 0.6793926564221918,
+      "learning_rate": 1.9538534690356225e-06,
+      "loss": 0.014,
+      "step": 3611
+    },
+    {
+      "epoch": 0.7174495977753501,
+      "grad_norm": 0.5600807419635767,
+      "learning_rate": 1.951302620543268e-06,
+      "loss": 0.0142,
+      "step": 3612
+    },
+    {
+      "epoch": 0.7176482272320985,
+      "grad_norm": 0.4854484281401081,
+      "learning_rate": 1.9487530345134464e-06,
+      "loss": 0.01,
+      "step": 3613
+    },
+    {
+      "epoch": 0.717846856688847,
+      "grad_norm": 0.9611540312341723,
+      "learning_rate": 1.946204712001936e-06,
+      "loss": 0.0098,
+      "step": 3614
+    },
+    {
+      "epoch": 0.7180454861455954,
+      "grad_norm": 0.5517679891572496,
+      "learning_rate": 1.9436576540639985e-06,
+      "loss": 0.0096,
+      "step": 3615
+    },
+    {
+      "epoch": 0.7182441156023438,
+      "grad_norm": 0.41701208366022036,
+      "learning_rate": 1.9411118617543634e-06,
+      "loss": 0.0077,
+      "step": 3616
+    },
+    {
+      "epoch": 0.7184427450590922,
+      "grad_norm": 0.36515434166067073,
+      "learning_rate": 1.938567336127245e-06,
+      "loss": 0.0145,
+      "step": 3617
+    },
+    {
+      "epoch": 0.7186413745158406,
+      "grad_norm": 0.42121979670363874,
+      "learning_rate": 1.936024078236325e-06,
+      "loss": 0.0089,
+      "step": 3618
+    },
+    {
+      "epoch": 0.7188400039725892,
+      "grad_norm": 0.43493645872145054,
+      "learning_rate": 1.9334820891347663e-06,
+      "loss": 0.0106,
+      "step": 3619
+    },
+    {
+      "epoch": 0.7190386334293376,
+      "grad_norm": 0.39201007950116096,
+      "learning_rate": 1.930941369875205e-06,
+      "loss": 0.0066,
+      "step": 3620
+    },
+    {
+      "epoch": 0.719237262886086,
+      "grad_norm": 0.44616612890490304,
+      "learning_rate": 1.928401921509747e-06,
+      "loss": 0.0089,
+      "step": 3621
+    },
+    {
+      "epoch": 0.7194358923428344,
+      "grad_norm": 0.5005235650718418,
+      "learning_rate": 1.9258637450899796e-06,
+      "loss": 0.0082,
+      "step": 3622
+    },
+    {
+      "epoch": 0.7196345217995829,
+      "grad_norm": 0.9873137170116204,
+      "learning_rate": 1.9233268416669547e-06,
+      "loss": 0.0143,
+      "step": 3623
+    },
+    {
+      "epoch": 0.7198331512563313,
+      "grad_norm": 0.9359450207876253,
+      "learning_rate": 1.920791212291206e-06,
+      "loss": 0.0142,
+      "step": 3624
+    },
+    {
+      "epoch": 0.7200317807130797,
+      "grad_norm": 0.8723213929265855,
+      "learning_rate": 1.9182568580127304e-06,
+      "loss": 0.0147,
+      "step": 3625
+    },
+    {
+      "epoch": 0.7202304101698281,
+      "grad_norm": 0.7238675157448663,
+      "learning_rate": 1.9157237798810037e-06,
+      "loss": 0.0116,
+      "step": 3626
+    },
+    {
+      "epoch": 0.7204290396265767,
+      "grad_norm": 0.8076205888225351,
+      "learning_rate": 1.9131919789449733e-06,
+      "loss": 0.0177,
+      "step": 3627
+    },
+    {
+      "epoch": 0.7206276690833251,
+      "grad_norm": 1.0779343741041851,
+      "learning_rate": 1.910661456253051e-06,
+      "loss": 0.021,
+      "step": 3628
+    },
+    {
+      "epoch": 0.7208262985400735,
+      "grad_norm": 0.5096320669800932,
+      "learning_rate": 1.9081322128531277e-06,
+      "loss": 0.0114,
+      "step": 3629
+    },
+    {
+      "epoch": 0.7210249279968219,
+      "grad_norm": 0.6623908050965808,
+      "learning_rate": 1.905604249792557e-06,
+      "loss": 0.0052,
+      "step": 3630
+    },
+    {
+      "epoch": 0.7212235574535704,
+      "grad_norm": 0.47990832396620436,
+      "learning_rate": 1.9030775681181696e-06,
+      "loss": 0.0091,
+      "step": 3631
+    },
+    {
+      "epoch": 0.7214221869103188,
+      "grad_norm": 0.7334184988391341,
+      "learning_rate": 1.9005521688762585e-06,
+      "loss": 0.0159,
+      "step": 3632
+    },
+    {
+      "epoch": 0.7216208163670672,
+      "grad_norm": 0.4980418429190839,
+      "learning_rate": 1.89802805311259e-06,
+      "loss": 0.0128,
+      "step": 3633
+    },
+    {
+      "epoch": 0.7218194458238156,
+      "grad_norm": 0.47426037363700535,
+      "learning_rate": 1.8955052218724002e-06,
+      "loss": 0.0085,
+      "step": 3634
+    },
+    {
+      "epoch": 0.722018075280564,
+      "grad_norm": 0.5843280660499018,
+      "learning_rate": 1.892983676200389e-06,
+      "loss": 0.0122,
+      "step": 3635
+    },
+    {
+      "epoch": 0.7222167047373126,
+      "grad_norm": 0.8246276543714378,
+      "learning_rate": 1.8904634171407238e-06,
+      "loss": 0.0095,
+      "step": 3636
+    },
+    {
+      "epoch": 0.722415334194061,
+      "grad_norm": 1.1929203320268906,
+      "learning_rate": 1.887944445737046e-06,
+      "loss": 0.0182,
+      "step": 3637
+    },
+    {
+      "epoch": 0.7226139636508094,
+      "grad_norm": 0.46024686499789297,
+      "learning_rate": 1.8854267630324547e-06,
+      "loss": 0.0061,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7228125931075579,
+      "grad_norm": 0.3041353357374024,
+      "learning_rate": 1.882910370069524e-06,
+      "loss": 0.006,
+      "step": 3639
+    },
+    {
+      "epoch": 0.7230112225643063,
+      "grad_norm": 0.566018431463613,
+      "learning_rate": 1.8803952678902853e-06,
+      "loss": 0.0125,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7232098520210547,
+      "grad_norm": 0.3629859525601858,
+      "learning_rate": 1.877881457536244e-06,
+      "loss": 0.0135,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7234084814778031,
+      "grad_norm": 0.49111992542952826,
+      "learning_rate": 1.8753689400483627e-06,
+      "loss": 0.0075,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7236071109345515,
+      "grad_norm": 0.3138065124597806,
+      "learning_rate": 1.8728577164670747e-06,
+      "loss": 0.0043,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7238057403913001,
+      "grad_norm": 0.9114034734955218,
+      "learning_rate": 1.8703477878322763e-06,
+      "loss": 0.0131,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7240043698480485,
+      "grad_norm": 0.4633138637914843,
+      "learning_rate": 1.8678391551833225e-06,
+      "loss": 0.0076,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7242029993047969,
+      "grad_norm": 0.5559865923945981,
+      "learning_rate": 1.8653318195590403e-06,
+      "loss": 0.0102,
+      "step": 3646
+    },
+    {
+      "epoch": 0.7244016287615453,
+      "grad_norm": 0.45909590775240494,
+      "learning_rate": 1.8628257819977102e-06,
+      "loss": 0.0114,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7246002582182938,
+      "grad_norm": 0.5720442731394239,
+      "learning_rate": 1.8603210435370845e-06,
+      "loss": 0.0139,
+      "step": 3648
+    },
+    {
+      "epoch": 0.7247988876750422,
+      "grad_norm": 1.0088288086073944,
+      "learning_rate": 1.8578176052143682e-06,
+      "loss": 0.0105,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7249975171317906,
+      "grad_norm": 0.341268216037555,
+      "learning_rate": 1.8553154680662355e-06,
+      "loss": 0.0102,
+      "step": 3650
+    },
+    {
+      "epoch": 0.725196146588539,
+      "grad_norm": 0.5430532731720251,
+      "learning_rate": 1.8528146331288198e-06,
+      "loss": 0.0056,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7253947760452876,
+      "grad_norm": 0.6547103394450872,
+      "learning_rate": 1.8503151014377108e-06,
+      "loss": 0.0141,
+      "step": 3652
+    },
+    {
+      "epoch": 0.725593405502036,
+      "grad_norm": 0.7543034404187979,
+      "learning_rate": 1.8478168740279662e-06,
+      "loss": 0.0146,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7257920349587844,
+      "grad_norm": 0.8527254403878096,
+      "learning_rate": 1.8453199519340959e-06,
+      "loss": 0.0108,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7259906644155328,
+      "grad_norm": 0.8087062079395743,
+      "learning_rate": 1.8428243361900754e-06,
+      "loss": 0.0127,
+      "step": 3655
+    },
+    {
+      "epoch": 0.7261892938722813,
+      "grad_norm": 0.906253089444188,
+      "learning_rate": 1.840330027829334e-06,
+      "loss": 0.0245,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7263879233290297,
+      "grad_norm": 0.6648053391954097,
+      "learning_rate": 1.8378370278847646e-06,
+      "loss": 0.012,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7265865527857781,
+      "grad_norm": 0.8741980778803626,
+      "learning_rate": 1.835345337388718e-06,
+      "loss": 0.0136,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7267851822425265,
+      "grad_norm": 0.637411767418505,
+      "learning_rate": 1.8328549573729948e-06,
+      "loss": 0.0148,
+      "step": 3659
+    },
+    {
+      "epoch": 0.726983811699275,
+      "grad_norm": 0.6026345033209373,
+      "learning_rate": 1.8303658888688642e-06,
+      "loss": 0.011,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7271824411560235,
+      "grad_norm": 0.7594705123411722,
+      "learning_rate": 1.8278781329070422e-06,
+      "loss": 0.017,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7273810706127719,
+      "grad_norm": 0.41929886968876007,
+      "learning_rate": 1.8253916905177093e-06,
+      "loss": 0.0134,
+      "step": 3662
+    },
+    {
+      "epoch": 0.7275797000695203,
+      "grad_norm": 1.0949746883385296,
+      "learning_rate": 1.8229065627305003e-06,
+      "loss": 0.0237,
+      "step": 3663
+    },
+    {
+      "epoch": 0.7277783295262688,
+      "grad_norm": 0.3118311043232577,
+      "learning_rate": 1.8204227505744998e-06,
+      "loss": 0.0068,
+      "step": 3664
+    },
+    {
+      "epoch": 0.7279769589830172,
+      "grad_norm": 0.659409007129231,
+      "learning_rate": 1.817940255078256e-06,
+      "loss": 0.01,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7281755884397656,
+      "grad_norm": 0.5934643848129301,
+      "learning_rate": 1.8154590772697644e-06,
+      "loss": 0.0096,
+      "step": 3666
+    },
+    {
+      "epoch": 0.728374217896514,
+      "grad_norm": 0.4976392266560354,
+      "learning_rate": 1.8129792181764817e-06,
+      "loss": 0.013,
+      "step": 3667
+    },
+    {
+      "epoch": 0.7285728473532624,
+      "grad_norm": 0.47726296885447955,
+      "learning_rate": 1.810500678825311e-06,
+      "loss": 0.0144,
+      "step": 3668
+    },
+    {
+      "epoch": 0.728771476810011,
+      "grad_norm": 0.22945888004342208,
+      "learning_rate": 1.8080234602426155e-06,
+      "loss": 0.0047,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7289701062667594,
+      "grad_norm": 0.5110342276655542,
+      "learning_rate": 1.8055475634542102e-06,
+      "loss": 0.0078,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7291687357235078,
+      "grad_norm": 0.31793929875854604,
+      "learning_rate": 1.8030729894853582e-06,
+      "loss": 0.0134,
+      "step": 3671
+    },
+    {
+      "epoch": 0.7293673651802562,
+      "grad_norm": 0.7487761059663722,
+      "learning_rate": 1.8005997393607816e-06,
+      "loss": 0.0206,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7295659946370047,
+      "grad_norm": 0.2538735229220153,
+      "learning_rate": 1.7981278141046472e-06,
+      "loss": 0.0088,
+      "step": 3673
+    },
+    {
+      "epoch": 0.7297646240937531,
+      "grad_norm": 0.4930426210508408,
+      "learning_rate": 1.7956572147405798e-06,
+      "loss": 0.0116,
+      "step": 3674
+    },
+    {
+      "epoch": 0.7299632535505015,
+      "grad_norm": 0.46080199402940236,
+      "learning_rate": 1.7931879422916499e-06,
+      "loss": 0.0054,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7301618830072499,
+      "grad_norm": 0.44044922022937444,
+      "learning_rate": 1.7907199977803809e-06,
+      "loss": 0.0099,
+      "step": 3676
+    },
+    {
+      "epoch": 0.7303605124639984,
+      "grad_norm": 0.42505852730210714,
+      "learning_rate": 1.788253382228749e-06,
+      "loss": 0.0196,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7305591419207469,
+      "grad_norm": 0.3898875170650358,
+      "learning_rate": 1.7857880966581738e-06,
+      "loss": 0.0051,
+      "step": 3678
+    },
+    {
+      "epoch": 0.7307577713774953,
+      "grad_norm": 0.49491000937059526,
+      "learning_rate": 1.7833241420895304e-06,
+      "loss": 0.011,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7309564008342437,
+      "grad_norm": 0.8767113299077823,
+      "learning_rate": 1.7808615195431366e-06,
+      "loss": 0.0131,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7311550302909922,
+      "grad_norm": 0.6506677708061251,
+      "learning_rate": 1.778400230038766e-06,
+      "loss": 0.0141,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7313536597477406,
+      "grad_norm": 0.31762522470848703,
+      "learning_rate": 1.775940274595634e-06,
+      "loss": 0.0072,
+      "step": 3682
+    },
+    {
+      "epoch": 0.731552289204489,
+      "grad_norm": 0.3400734480718243,
+      "learning_rate": 1.7734816542324034e-06,
+      "loss": 0.0069,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7317509186612374,
+      "grad_norm": 0.3523223550612423,
+      "learning_rate": 1.7710243699671908e-06,
+      "loss": 0.0089,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7319495481179858,
+      "grad_norm": 0.26978468960487634,
+      "learning_rate": 1.768568422817551e-06,
+      "loss": 0.0061,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7321481775747344,
+      "grad_norm": 0.5668983473598859,
+      "learning_rate": 1.7661138138004918e-06,
+      "loss": 0.0153,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7323468070314828,
+      "grad_norm": 1.5335058937602353,
+      "learning_rate": 1.763660543932465e-06,
+      "loss": 0.0177,
+      "step": 3687
+    },
+    {
+      "epoch": 0.7325454364882312,
+      "grad_norm": 0.5330209491574611,
+      "learning_rate": 1.7612086142293643e-06,
+      "loss": 0.0097,
+      "step": 3688
+    },
+    {
+      "epoch": 0.7327440659449797,
+      "grad_norm": 0.5463971064029184,
+      "learning_rate": 1.758758025706535e-06,
+      "loss": 0.0125,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7329426954017281,
+      "grad_norm": 0.4545424691631491,
+      "learning_rate": 1.7563087793787597e-06,
+      "loss": 0.0095,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7331413248584765,
+      "grad_norm": 0.4004144836163272,
+      "learning_rate": 1.753860876260272e-06,
+      "loss": 0.0109,
+      "step": 3691
+    },
+    {
+      "epoch": 0.7333399543152249,
+      "grad_norm": 0.4843018902925179,
+      "learning_rate": 1.751414317364743e-06,
+      "loss": 0.0128,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7335385837719733,
+      "grad_norm": 0.27511747492039373,
+      "learning_rate": 1.7489691037052925e-06,
+      "loss": 0.0068,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7337372132287219,
+      "grad_norm": 0.36643270106143355,
+      "learning_rate": 1.7465252362944818e-06,
+      "loss": 0.0093,
+      "step": 3694
+    },
+    {
+      "epoch": 0.7339358426854703,
+      "grad_norm": 0.320728805385628,
+      "learning_rate": 1.7440827161443108e-06,
+      "loss": 0.0115,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7341344721422187,
+      "grad_norm": 0.42125602343146334,
+      "learning_rate": 1.7416415442662283e-06,
+      "loss": 0.0135,
+      "step": 3696
+    },
+    {
+      "epoch": 0.7343331015989671,
+      "grad_norm": 0.418662760132342,
+      "learning_rate": 1.7392017216711178e-06,
+      "loss": 0.0113,
+      "step": 3697
+    },
+    {
+      "epoch": 0.7345317310557156,
+      "grad_norm": 0.29967942764047406,
+      "learning_rate": 1.7367632493693098e-06,
+      "loss": 0.0108,
+      "step": 3698
+    },
+    {
+      "epoch": 0.734730360512464,
+      "grad_norm": 0.683085498593672,
+      "learning_rate": 1.7343261283705714e-06,
+      "loss": 0.0108,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7349289899692124,
+      "grad_norm": 0.3669500417485653,
+      "learning_rate": 1.731890359684113e-06,
+      "loss": 0.0086,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7351276194259608,
+      "grad_norm": 0.7989281415130531,
+      "learning_rate": 1.7294559443185854e-06,
+      "loss": 0.0114,
+      "step": 3701
+    },
+    {
+      "epoch": 0.7353262488827093,
+      "grad_norm": 0.3908467521711616,
+      "learning_rate": 1.727022883282074e-06,
+      "loss": 0.0079,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7355248783394578,
+      "grad_norm": 0.4811819714857938,
+      "learning_rate": 1.7245911775821112e-06,
+      "loss": 0.005,
+      "step": 3703
+    },
+    {
+      "epoch": 0.7357235077962062,
+      "grad_norm": 0.5482194796559414,
+      "learning_rate": 1.7221608282256597e-06,
+      "loss": 0.0088,
+      "step": 3704
+    },
+    {
+      "epoch": 0.7359221372529546,
+      "grad_norm": 0.3769237988240613,
+      "learning_rate": 1.7197318362191284e-06,
+      "loss": 0.0138,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7361207667097031,
+      "grad_norm": 0.6344995974195727,
+      "learning_rate": 1.717304202568359e-06,
+      "loss": 0.0169,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7363193961664515,
+      "grad_norm": 0.5850068173437899,
+      "learning_rate": 1.7148779282786305e-06,
+      "loss": 0.0138,
+      "step": 3707
+    },
+    {
+      "epoch": 0.7365180256231999,
+      "grad_norm": 0.7795863924661017,
+      "learning_rate": 1.712453014354663e-06,
+      "loss": 0.0122,
+      "step": 3708
+    },
+    {
+      "epoch": 0.7367166550799483,
+      "grad_norm": 0.35280976439404965,
+      "learning_rate": 1.7100294618006092e-06,
+      "loss": 0.0084,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7369152845366967,
+      "grad_norm": 0.5700366010660607,
+      "learning_rate": 1.7076072716200616e-06,
+      "loss": 0.0155,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7371139139934453,
+      "grad_norm": 0.40895343575304466,
+      "learning_rate": 1.7051864448160444e-06,
+      "loss": 0.0084,
+      "step": 3711
+    },
+    {
+      "epoch": 0.7373125434501937,
+      "grad_norm": 0.4305656202174194,
+      "learning_rate": 1.7027669823910208e-06,
+      "loss": 0.0111,
+      "step": 3712
+    },
+    {
+      "epoch": 0.7375111729069421,
+      "grad_norm": 0.984538631213734,
+      "learning_rate": 1.7003488853468897e-06,
+      "loss": 0.0131,
+      "step": 3713
+    },
+    {
+      "epoch": 0.7377098023636905,
+      "grad_norm": 0.5010493377039495,
+      "learning_rate": 1.6979321546849786e-06,
+      "loss": 0.0063,
+      "step": 3714
+    },
+    {
+      "epoch": 0.737908431820439,
+      "grad_norm": 0.45048026983636796,
+      "learning_rate": 1.6955167914060578e-06,
+      "loss": 0.012,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7381070612771874,
+      "grad_norm": 0.2741879991054503,
+      "learning_rate": 1.6931027965103224e-06,
+      "loss": 0.0057,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7383056907339358,
+      "grad_norm": 0.4130916332257633,
+      "learning_rate": 1.6906901709974093e-06,
+      "loss": 0.0115,
+      "step": 3717
+    },
+    {
+      "epoch": 0.7385043201906842,
+      "grad_norm": 0.561420411364561,
+      "learning_rate": 1.6882789158663803e-06,
+      "loss": 0.0128,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7387029496474327,
+      "grad_norm": 0.22467074037808235,
+      "learning_rate": 1.6858690321157362e-06,
+      "loss": 0.0036,
+      "step": 3719
+    },
+    {
+      "epoch": 0.7389015791041812,
+      "grad_norm": 0.7303477211037094,
+      "learning_rate": 1.6834605207434084e-06,
+      "loss": 0.0167,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7391002085609296,
+      "grad_norm": 0.5588024743579839,
+      "learning_rate": 1.6810533827467563e-06,
+      "loss": 0.0104,
+      "step": 3721
+    },
+    {
+      "epoch": 0.739298838017678,
+      "grad_norm": 0.2374237543068285,
+      "learning_rate": 1.6786476191225764e-06,
+      "loss": 0.0045,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7394974674744265,
+      "grad_norm": 0.4732242064561535,
+      "learning_rate": 1.6762432308670895e-06,
+      "loss": 0.0095,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7396960969311749,
+      "grad_norm": 1.3353741885909276,
+      "learning_rate": 1.6738402189759539e-06,
+      "loss": 0.0084,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7398947263879233,
+      "grad_norm": 0.7402268023476685,
+      "learning_rate": 1.671438584444251e-06,
+      "loss": 0.0161,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7400933558446717,
+      "grad_norm": 0.48374235673574467,
+      "learning_rate": 1.6690383282664975e-06,
+      "loss": 0.0117,
+      "step": 3726
+    },
+    {
+      "epoch": 0.7402919853014202,
+      "grad_norm": 0.3143118173710068,
+      "learning_rate": 1.6666394514366374e-06,
+      "loss": 0.0106,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7404906147581687,
+      "grad_norm": 0.622043363621592,
+      "learning_rate": 1.6642419549480414e-06,
+      "loss": 0.0069,
+      "step": 3728
+    },
+    {
+      "epoch": 0.7406892442149171,
+      "grad_norm": 0.5599094194736531,
+      "learning_rate": 1.6618458397935128e-06,
+      "loss": 0.015,
+      "step": 3729
+    },
+    {
+      "epoch": 0.7408878736716655,
+      "grad_norm": 0.48331856066211576,
+      "learning_rate": 1.6594511069652786e-06,
+      "loss": 0.0066,
+      "step": 3730
+    },
+    {
+      "epoch": 0.741086503128414,
+      "grad_norm": 0.34166598414346067,
+      "learning_rate": 1.6570577574549945e-06,
+      "loss": 0.0066,
+      "step": 3731
+    },
+    {
+      "epoch": 0.7412851325851624,
+      "grad_norm": 0.427571456888473,
+      "learning_rate": 1.6546657922537467e-06,
+      "loss": 0.0069,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7414837620419108,
+      "grad_norm": 0.3951042263387438,
+      "learning_rate": 1.6522752123520431e-06,
+      "loss": 0.0057,
+      "step": 3733
+    },
+    {
+      "epoch": 0.7416823914986592,
+      "grad_norm": 0.5078153456049683,
+      "learning_rate": 1.6498860187398225e-06,
+      "loss": 0.0113,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7418810209554076,
+      "grad_norm": 0.8021479414729211,
+      "learning_rate": 1.6474982124064453e-06,
+      "loss": 0.0165,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7420796504121562,
+      "grad_norm": 0.2070717809242952,
+      "learning_rate": 1.6451117943407014e-06,
+      "loss": 0.0068,
+      "step": 3736
+    },
+    {
+      "epoch": 0.7422782798689046,
+      "grad_norm": 1.1209489750991877,
+      "learning_rate": 1.6427267655308048e-06,
+      "loss": 0.0134,
+      "step": 3737
+    },
+    {
+      "epoch": 0.742476909325653,
+      "grad_norm": 0.9481691520290123,
+      "learning_rate": 1.6403431269643916e-06,
+      "loss": 0.0169,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7426755387824014,
+      "grad_norm": 0.6585340668755794,
+      "learning_rate": 1.6379608796285263e-06,
+      "loss": 0.01,
+      "step": 3739
+    },
+    {
+      "epoch": 0.7428741682391499,
+      "grad_norm": 0.7454877441367229,
+      "learning_rate": 1.635580024509692e-06,
+      "loss": 0.0084,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7430727976958983,
+      "grad_norm": 0.6830037620298528,
+      "learning_rate": 1.6332005625938025e-06,
+      "loss": 0.0139,
+      "step": 3741
+    },
+    {
+      "epoch": 0.7432714271526467,
+      "grad_norm": 0.40026913724686264,
+      "learning_rate": 1.6308224948661867e-06,
+      "loss": 0.0102,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7434700566093951,
+      "grad_norm": 1.211034939945853,
+      "learning_rate": 1.6284458223116011e-06,
+      "loss": 0.0208,
+      "step": 3743
+    },
+    {
+      "epoch": 0.7436686860661436,
+      "grad_norm": 0.4950643350049753,
+      "learning_rate": 1.6260705459142268e-06,
+      "loss": 0.0099,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7438673155228921,
+      "grad_norm": 0.7561528087411485,
+      "learning_rate": 1.6236966666576586e-06,
+      "loss": 0.015,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7440659449796405,
+      "grad_norm": 0.4587296244272926,
+      "learning_rate": 1.6213241855249211e-06,
+      "loss": 0.0105,
+      "step": 3746
+    },
+    {
+      "epoch": 0.7442645744363889,
+      "grad_norm": 0.35825062637065686,
+      "learning_rate": 1.6189531034984534e-06,
+      "loss": 0.0071,
+      "step": 3747
+    },
+    {
+      "epoch": 0.7444632038931374,
+      "grad_norm": 0.1667528612749447,
+      "learning_rate": 1.616583421560121e-06,
+      "loss": 0.0034,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7446618333498858,
+      "grad_norm": 1.4414518890486638,
+      "learning_rate": 1.6142151406912043e-06,
+      "loss": 0.0252,
+      "step": 3749
+    },
+    {
+      "epoch": 0.7448604628066342,
+      "grad_norm": 0.8832252217740378,
+      "learning_rate": 1.6118482618724073e-06,
+      "loss": 0.0266,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7450590922633826,
+      "grad_norm": 0.76220401745124,
+      "learning_rate": 1.6094827860838535e-06,
+      "loss": 0.0106,
+      "step": 3751
+    },
+    {
+      "epoch": 0.745257721720131,
+      "grad_norm": 0.9771556397073955,
+      "learning_rate": 1.6071187143050809e-06,
+      "loss": 0.0122,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7454563511768796,
+      "grad_norm": 0.3409011178631712,
+      "learning_rate": 1.6047560475150532e-06,
+      "loss": 0.0044,
+      "step": 3753
+    },
+    {
+      "epoch": 0.745654980633628,
+      "grad_norm": 1.2118712960840325,
+      "learning_rate": 1.6023947866921452e-06,
+      "loss": 0.0163,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7458536100903764,
+      "grad_norm": 0.7180041710450296,
+      "learning_rate": 1.6000349328141528e-06,
+      "loss": 0.0171,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7460522395471249,
+      "grad_norm": 0.8759814111147064,
+      "learning_rate": 1.597676486858291e-06,
+      "loss": 0.0129,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7462508690038733,
+      "grad_norm": 0.3860414331142917,
+      "learning_rate": 1.5953194498011876e-06,
+      "loss": 0.0081,
+      "step": 3757
+    },
+    {
+      "epoch": 0.7464494984606217,
+      "grad_norm": 0.6321819223523495,
+      "learning_rate": 1.5929638226188915e-06,
+      "loss": 0.0144,
+      "step": 3758
+    },
+    {
+      "epoch": 0.7466481279173701,
+      "grad_norm": 0.21545187521317896,
+      "learning_rate": 1.5906096062868638e-06,
+      "loss": 0.0035,
+      "step": 3759
+    },
+    {
+      "epoch": 0.7468467573741185,
+      "grad_norm": 0.47253677184045745,
+      "learning_rate": 1.5882568017799848e-06,
+      "loss": 0.0099,
+      "step": 3760
+    },
+    {
+      "epoch": 0.747045386830867,
+      "grad_norm": 0.678212792886403,
+      "learning_rate": 1.5859054100725463e-06,
+      "loss": 0.0164,
+      "step": 3761
+    },
+    {
+      "epoch": 0.7472440162876155,
+      "grad_norm": 0.5852719573661888,
+      "learning_rate": 1.5835554321382585e-06,
+      "loss": 0.0111,
+      "step": 3762
+    },
+    {
+      "epoch": 0.7474426457443639,
+      "grad_norm": 0.46344554315419095,
+      "learning_rate": 1.5812068689502464e-06,
+      "loss": 0.0174,
+      "step": 3763
+    },
+    {
+      "epoch": 0.7476412752011123,
+      "grad_norm": 0.6136914065626226,
+      "learning_rate": 1.5788597214810448e-06,
+      "loss": 0.0084,
+      "step": 3764
+    },
+    {
+      "epoch": 0.7478399046578608,
+      "grad_norm": 0.6281579179223541,
+      "learning_rate": 1.576513990702608e-06,
+      "loss": 0.0083,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7480385341146092,
+      "grad_norm": 0.543301875004366,
+      "learning_rate": 1.5741696775862975e-06,
+      "loss": 0.009,
+      "step": 3766
+    },
+    {
+      "epoch": 0.7482371635713576,
+      "grad_norm": 0.36572588305171494,
+      "learning_rate": 1.5718267831028944e-06,
+      "loss": 0.0119,
+      "step": 3767
+    },
+    {
+      "epoch": 0.748435793028106,
+      "grad_norm": 0.24068019330565446,
+      "learning_rate": 1.5694853082225859e-06,
+      "loss": 0.0055,
+      "step": 3768
+    },
+    {
+      "epoch": 0.7486344224848545,
+      "grad_norm": 0.41680849368459344,
+      "learning_rate": 1.5671452539149761e-06,
+      "loss": 0.0062,
+      "step": 3769
+    },
+    {
+      "epoch": 0.748833051941603,
+      "grad_norm": 0.8860868458069969,
+      "learning_rate": 1.5648066211490804e-06,
+      "loss": 0.0158,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7490316813983514,
+      "grad_norm": 0.36239580091258583,
+      "learning_rate": 1.5624694108933208e-06,
+      "loss": 0.0079,
+      "step": 3771
+    },
+    {
+      "epoch": 0.7492303108550998,
+      "grad_norm": 0.855233140426005,
+      "learning_rate": 1.560133624115538e-06,
+      "loss": 0.0133,
+      "step": 3772
+    },
+    {
+      "epoch": 0.7494289403118483,
+      "grad_norm": 0.5163747324216429,
+      "learning_rate": 1.5577992617829745e-06,
+      "loss": 0.0119,
+      "step": 3773
+    },
+    {
+      "epoch": 0.7496275697685967,
+      "grad_norm": 4.943484511480794,
+      "learning_rate": 1.5554663248622914e-06,
+      "loss": 0.0085,
+      "step": 3774
+    },
+    {
+      "epoch": 0.7498261992253451,
+      "grad_norm": 0.6847981517474558,
+      "learning_rate": 1.5531348143195524e-06,
+      "loss": 0.0158,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7500248286820935,
+      "grad_norm": 0.24726926855367248,
+      "learning_rate": 1.5508047311202346e-06,
+      "loss": 0.0063,
+      "step": 3776
+    },
+    {
+      "epoch": 0.750223458138842,
+      "grad_norm": 1.0653880290235649,
+      "learning_rate": 1.548476076229225e-06,
+      "loss": 0.0152,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7504220875955905,
+      "grad_norm": 0.41225668837673585,
+      "learning_rate": 1.5461488506108163e-06,
+      "loss": 0.0079,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7506207170523389,
+      "grad_norm": 0.39688034716378284,
+      "learning_rate": 1.5438230552287076e-06,
+      "loss": 0.0133,
+      "step": 3779
+    },
+    {
+      "epoch": 0.7508193465090873,
+      "grad_norm": 0.7160254723600693,
+      "learning_rate": 1.5414986910460127e-06,
+      "loss": 0.0119,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7510179759658357,
+      "grad_norm": 0.5240990626907865,
+      "learning_rate": 1.5391757590252448e-06,
+      "loss": 0.0115,
+      "step": 3781
+    },
+    {
+      "epoch": 0.7512166054225842,
+      "grad_norm": 1.2549979086012635,
+      "learning_rate": 1.536854260128331e-06,
+      "loss": 0.0133,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7514152348793326,
+      "grad_norm": 0.9872465010490661,
+      "learning_rate": 1.5345341953165982e-06,
+      "loss": 0.0137,
+      "step": 3783
+    },
+    {
+      "epoch": 0.751613864336081,
+      "grad_norm": 0.34007165388302485,
+      "learning_rate": 1.5322155655507859e-06,
+      "loss": 0.0084,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7518124937928294,
+      "grad_norm": 0.494044558950559,
+      "learning_rate": 1.5298983717910342e-06,
+      "loss": 0.0127,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7520111232495779,
+      "grad_norm": 0.8210240909130295,
+      "learning_rate": 1.5275826149968913e-06,
+      "loss": 0.0089,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7522097527063264,
+      "grad_norm": 0.47775536674059704,
+      "learning_rate": 1.5252682961273125e-06,
+      "loss": 0.0107,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7524083821630748,
+      "grad_norm": 0.42651256912812496,
+      "learning_rate": 1.5229554161406502e-06,
+      "loss": 0.0104,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7526070116198232,
+      "grad_norm": 0.5814023196245535,
+      "learning_rate": 1.52064397599467e-06,
+      "loss": 0.0115,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7528056410765717,
+      "grad_norm": 0.39860990655481554,
+      "learning_rate": 1.5183339766465332e-06,
+      "loss": 0.0108,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7530042705333201,
+      "grad_norm": 0.3816121403616372,
+      "learning_rate": 1.5160254190528118e-06,
+      "loss": 0.0073,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7532028999900685,
+      "grad_norm": 0.5562129339192128,
+      "learning_rate": 1.5137183041694736e-06,
+      "loss": 0.0145,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7534015294468169,
+      "grad_norm": 0.35565917270513564,
+      "learning_rate": 1.5114126329518947e-06,
+      "loss": 0.0088,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7536001589035654,
+      "grad_norm": 0.47105502946267286,
+      "learning_rate": 1.5091084063548527e-06,
+      "loss": 0.0126,
+      "step": 3794
+    },
+    {
+      "epoch": 0.7537987883603139,
+      "grad_norm": 0.6185552661380914,
+      "learning_rate": 1.506805625332522e-06,
+      "loss": 0.0123,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7539974178170623,
+      "grad_norm": 0.40215289956017086,
+      "learning_rate": 1.504504290838485e-06,
+      "loss": 0.0075,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7541960472738107,
+      "grad_norm": 0.4782485953056544,
+      "learning_rate": 1.5022044038257195e-06,
+      "loss": 0.0162,
+      "step": 3797
+    },
+    {
+      "epoch": 0.7543946767305592,
+      "grad_norm": 0.7324829878695893,
+      "learning_rate": 1.4999059652466085e-06,
+      "loss": 0.0105,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7545933061873076,
+      "grad_norm": 0.47495420804446264,
+      "learning_rate": 1.4976089760529311e-06,
+      "loss": 0.0139,
+      "step": 3799
+    },
+    {
+      "epoch": 0.754791935644056,
+      "grad_norm": 0.30470882134001215,
+      "learning_rate": 1.4953134371958694e-06,
+      "loss": 0.0044,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7549905651008044,
+      "grad_norm": 0.5558913333675957,
+      "learning_rate": 1.4930193496260053e-06,
+      "loss": 0.0136,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7551891945575528,
+      "grad_norm": 0.5677602935945104,
+      "learning_rate": 1.4907267142933162e-06,
+      "loss": 0.007,
+      "step": 3802
+    },
+    {
+      "epoch": 0.7553878240143013,
+      "grad_norm": 0.44674486843299116,
+      "learning_rate": 1.488435532147181e-06,
+      "loss": 0.0097,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7555864534710498,
+      "grad_norm": 0.24979032118916042,
+      "learning_rate": 1.4861458041363736e-06,
+      "loss": 0.0067,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7557850829277982,
+      "grad_norm": 0.644903890844482,
+      "learning_rate": 1.4838575312090692e-06,
+      "loss": 0.0176,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7559837123845466,
+      "grad_norm": 0.4250844938968267,
+      "learning_rate": 1.481570714312842e-06,
+      "loss": 0.0095,
+      "step": 3806
+    },
+    {
+      "epoch": 0.7561823418412951,
+      "grad_norm": 0.5450830869397256,
+      "learning_rate": 1.4792853543946572e-06,
+      "loss": 0.014,
+      "step": 3807
+    },
+    {
+      "epoch": 0.7563809712980435,
+      "grad_norm": 0.2125242305530338,
+      "learning_rate": 1.477001452400883e-06,
+      "loss": 0.0067,
+      "step": 3808
+    },
+    {
+      "epoch": 0.7565796007547919,
+      "grad_norm": 0.6807746018493928,
+      "learning_rate": 1.4747190092772774e-06,
+      "loss": 0.0151,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7567782302115403,
+      "grad_norm": 0.42307268538948456,
+      "learning_rate": 1.4724380259690013e-06,
+      "loss": 0.0119,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7569768596682888,
+      "grad_norm": 0.6930123507211039,
+      "learning_rate": 1.470158503420605e-06,
+      "loss": 0.0104,
+      "step": 3811
+    },
+    {
+      "epoch": 0.7571754891250373,
+      "grad_norm": 0.3363804661976727,
+      "learning_rate": 1.4678804425760374e-06,
+      "loss": 0.0049,
+      "step": 3812
+    },
+    {
+      "epoch": 0.7573741185817857,
+      "grad_norm": 0.5791696042890436,
+      "learning_rate": 1.4656038443786425e-06,
+      "loss": 0.0119,
+      "step": 3813
+    },
+    {
+      "epoch": 0.7575727480385341,
+      "grad_norm": 0.30695431877058926,
+      "learning_rate": 1.4633287097711552e-06,
+      "loss": 0.0062,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7577713774952826,
+      "grad_norm": 0.5269032914372845,
+      "learning_rate": 1.4610550396957085e-06,
+      "loss": 0.0175,
+      "step": 3815
+    },
+    {
+      "epoch": 0.757970006952031,
+      "grad_norm": 0.38075958392152365,
+      "learning_rate": 1.458782835093825e-06,
+      "loss": 0.008,
+      "step": 3816
+    },
+    {
+      "epoch": 0.7581686364087794,
+      "grad_norm": 0.3840340193367981,
+      "learning_rate": 1.4565120969064222e-06,
+      "loss": 0.0062,
+      "step": 3817
+    },
+    {
+      "epoch": 0.7583672658655278,
+      "grad_norm": 0.44995171484252383,
+      "learning_rate": 1.454242826073814e-06,
+      "loss": 0.0105,
+      "step": 3818
+    },
+    {
+      "epoch": 0.7585658953222763,
+      "grad_norm": 0.755410175506028,
+      "learning_rate": 1.451975023535699e-06,
+      "loss": 0.0151,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7587645247790247,
+      "grad_norm": 0.4925822601592495,
+      "learning_rate": 1.4497086902311746e-06,
+      "loss": 0.0118,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7589631542357732,
+      "grad_norm": 0.33772272051963276,
+      "learning_rate": 1.4474438270987257e-06,
+      "loss": 0.0062,
+      "step": 3821
+    },
+    {
+      "epoch": 0.7591617836925216,
+      "grad_norm": 0.8290174834097047,
+      "learning_rate": 1.445180435076231e-06,
+      "loss": 0.0136,
+      "step": 3822
+    },
+    {
+      "epoch": 0.75936041314927,
+      "grad_norm": 0.43343542171958976,
+      "learning_rate": 1.4429185151009573e-06,
+      "loss": 0.0092,
+      "step": 3823
+    },
+    {
+      "epoch": 0.7595590426060185,
+      "grad_norm": 0.775111328446171,
+      "learning_rate": 1.4406580681095639e-06,
+      "loss": 0.0181,
+      "step": 3824
+    },
+    {
+      "epoch": 0.7597576720627669,
+      "grad_norm": 0.39166179536426354,
+      "learning_rate": 1.4383990950381022e-06,
+      "loss": 0.0088,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7599563015195153,
+      "grad_norm": 0.4925837143073198,
+      "learning_rate": 1.4361415968220082e-06,
+      "loss": 0.0101,
+      "step": 3826
+    },
+    {
+      "epoch": 0.7601549309762637,
+      "grad_norm": 0.32696898445858813,
+      "learning_rate": 1.43388557439611e-06,
+      "loss": 0.0086,
+      "step": 3827
+    },
+    {
+      "epoch": 0.7603535604330122,
+      "grad_norm": 0.4958561019808758,
+      "learning_rate": 1.4316310286946228e-06,
+      "loss": 0.0137,
+      "step": 3828
+    },
+    {
+      "epoch": 0.7605521898897607,
+      "grad_norm": 0.4975604655384996,
+      "learning_rate": 1.4293779606511527e-06,
+      "loss": 0.0136,
+      "step": 3829
+    },
+    {
+      "epoch": 0.7607508193465091,
+      "grad_norm": 0.49887393821198955,
+      "learning_rate": 1.4271263711986954e-06,
+      "loss": 0.0103,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7609494488032575,
+      "grad_norm": 0.3683363611647419,
+      "learning_rate": 1.4248762612696282e-06,
+      "loss": 0.0063,
+      "step": 3831
+    },
+    {
+      "epoch": 0.761148078260006,
+      "grad_norm": 0.3079985582051433,
+      "learning_rate": 1.4226276317957228e-06,
+      "loss": 0.0125,
+      "step": 3832
+    },
+    {
+      "epoch": 0.7613467077167544,
+      "grad_norm": 0.4074575929705946,
+      "learning_rate": 1.4203804837081308e-06,
+      "loss": 0.0075,
+      "step": 3833
+    },
+    {
+      "epoch": 0.7615453371735028,
+      "grad_norm": 0.49220265205108876,
+      "learning_rate": 1.4181348179373972e-06,
+      "loss": 0.0127,
+      "step": 3834
+    },
+    {
+      "epoch": 0.7617439666302512,
+      "grad_norm": 0.2819678338268758,
+      "learning_rate": 1.4158906354134472e-06,
+      "loss": 0.0082,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7619425960869997,
+      "grad_norm": 0.5745246590432623,
+      "learning_rate": 1.413647937065596e-06,
+      "loss": 0.0149,
+      "step": 3836
+    },
+    {
+      "epoch": 0.7621412255437482,
+      "grad_norm": 0.6022591411514704,
+      "learning_rate": 1.4114067238225438e-06,
+      "loss": 0.017,
+      "step": 3837
+    },
+    {
+      "epoch": 0.7623398550004966,
+      "grad_norm": 0.5962439448569538,
+      "learning_rate": 1.4091669966123717e-06,
+      "loss": 0.0079,
+      "step": 3838
+    },
+    {
+      "epoch": 0.762538484457245,
+      "grad_norm": 0.48725897414015545,
+      "learning_rate": 1.4069287563625523e-06,
+      "loss": 0.0122,
+      "step": 3839
+    },
+    {
+      "epoch": 0.7627371139139935,
+      "grad_norm": 0.6291651873134785,
+      "learning_rate": 1.404692003999935e-06,
+      "loss": 0.0142,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7629357433707419,
+      "grad_norm": 0.5770793807297394,
+      "learning_rate": 1.4024567404507606e-06,
+      "loss": 0.0129,
+      "step": 3841
+    },
+    {
+      "epoch": 0.7631343728274903,
+      "grad_norm": 0.4801313035665573,
+      "learning_rate": 1.4002229666406454e-06,
+      "loss": 0.0127,
+      "step": 3842
+    },
+    {
+      "epoch": 0.7633330022842387,
+      "grad_norm": 0.5099357827368747,
+      "learning_rate": 1.3979906834945944e-06,
+      "loss": 0.0139,
+      "step": 3843
+    },
+    {
+      "epoch": 0.7635316317409871,
+      "grad_norm": 0.3833860818815375,
+      "learning_rate": 1.3957598919369958e-06,
+      "loss": 0.0099,
+      "step": 3844
+    },
+    {
+      "epoch": 0.7637302611977356,
+      "grad_norm": 0.41162753016444315,
+      "learning_rate": 1.3935305928916154e-06,
+      "loss": 0.0108,
+      "step": 3845
+    },
+    {
+      "epoch": 0.7639288906544841,
+      "grad_norm": 0.3078560285036505,
+      "learning_rate": 1.3913027872816064e-06,
+      "loss": 0.0095,
+      "step": 3846
+    },
+    {
+      "epoch": 0.7641275201112325,
+      "grad_norm": 0.37973724509719714,
+      "learning_rate": 1.3890764760294979e-06,
+      "loss": 0.0074,
+      "step": 3847
+    },
+    {
+      "epoch": 0.764326149567981,
+      "grad_norm": 0.4498229089715063,
+      "learning_rate": 1.3868516600572056e-06,
+      "loss": 0.0091,
+      "step": 3848
+    },
+    {
+      "epoch": 0.7645247790247294,
+      "grad_norm": 0.37444629015372083,
+      "learning_rate": 1.3846283402860216e-06,
+      "loss": 0.0099,
+      "step": 3849
+    },
+    {
+      "epoch": 0.7647234084814778,
+      "grad_norm": 0.5009045045676753,
+      "learning_rate": 1.3824065176366225e-06,
+      "loss": 0.0105,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7649220379382262,
+      "grad_norm": 0.4475657849726896,
+      "learning_rate": 1.3801861930290623e-06,
+      "loss": 0.0064,
+      "step": 3851
+    },
+    {
+      "epoch": 0.7651206673949746,
+      "grad_norm": 0.774374062344075,
+      "learning_rate": 1.377967367382772e-06,
+      "loss": 0.0182,
+      "step": 3852
+    },
+    {
+      "epoch": 0.7653192968517231,
+      "grad_norm": 0.4792793910315004,
+      "learning_rate": 1.3757500416165686e-06,
+      "loss": 0.012,
+      "step": 3853
+    },
+    {
+      "epoch": 0.7655179263084716,
+      "grad_norm": 0.6802435312134761,
+      "learning_rate": 1.3735342166486448e-06,
+      "loss": 0.0146,
+      "step": 3854
+    },
+    {
+      "epoch": 0.76571655576522,
+      "grad_norm": 0.5822617738126644,
+      "learning_rate": 1.3713198933965687e-06,
+      "loss": 0.0102,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7659151852219684,
+      "grad_norm": 0.3228251141085814,
+      "learning_rate": 1.3691070727772927e-06,
+      "loss": 0.0058,
+      "step": 3856
+    },
+    {
+      "epoch": 0.7661138146787169,
+      "grad_norm": 0.5602496268212196,
+      "learning_rate": 1.36689575570714e-06,
+      "loss": 0.0138,
+      "step": 3857
+    },
+    {
+      "epoch": 0.7663124441354653,
+      "grad_norm": 0.8410546969132571,
+      "learning_rate": 1.3646859431018178e-06,
+      "loss": 0.0168,
+      "step": 3858
+    },
+    {
+      "epoch": 0.7665110735922137,
+      "grad_norm": 0.49153277899098013,
+      "learning_rate": 1.3624776358764046e-06,
+      "loss": 0.0101,
+      "step": 3859
+    },
+    {
+      "epoch": 0.7667097030489621,
+      "grad_norm": 0.4718983704981794,
+      "learning_rate": 1.3602708349453603e-06,
+      "loss": 0.0146,
+      "step": 3860
+    },
+    {
+      "epoch": 0.7669083325057106,
+      "grad_norm": 0.46323476790072704,
+      "learning_rate": 1.3580655412225192e-06,
+      "loss": 0.0141,
+      "step": 3861
+    },
+    {
+      "epoch": 0.767106961962459,
+      "grad_norm": 0.72794578293661,
+      "learning_rate": 1.3558617556210891e-06,
+      "loss": 0.0146,
+      "step": 3862
+    },
+    {
+      "epoch": 0.7673055914192075,
+      "grad_norm": 0.36007802301261976,
+      "learning_rate": 1.3536594790536584e-06,
+      "loss": 0.0073,
+      "step": 3863
+    },
+    {
+      "epoch": 0.7675042208759559,
+      "grad_norm": 0.6871366174736734,
+      "learning_rate": 1.3514587124321842e-06,
+      "loss": 0.0154,
+      "step": 3864
+    },
+    {
+      "epoch": 0.7677028503327044,
+      "grad_norm": 0.77934118848991,
+      "learning_rate": 1.3492594566680052e-06,
+      "loss": 0.0133,
+      "step": 3865
+    },
+    {
+      "epoch": 0.7679014797894528,
+      "grad_norm": 0.31816704466209683,
+      "learning_rate": 1.3470617126718272e-06,
+      "loss": 0.0114,
+      "step": 3866
+    },
+    {
+      "epoch": 0.7681001092462012,
+      "grad_norm": 0.39457341298241466,
+      "learning_rate": 1.344865481353736e-06,
+      "loss": 0.0088,
+      "step": 3867
+    },
+    {
+      "epoch": 0.7682987387029496,
+      "grad_norm": 0.3878354688307015,
+      "learning_rate": 1.34267076362319e-06,
+      "loss": 0.0051,
+      "step": 3868
+    },
+    {
+      "epoch": 0.768497368159698,
+      "grad_norm": 0.5176469767858003,
+      "learning_rate": 1.3404775603890175e-06,
+      "loss": 0.0102,
+      "step": 3869
+    },
+    {
+      "epoch": 0.7686959976164465,
+      "grad_norm": 0.6546341810751645,
+      "learning_rate": 1.3382858725594233e-06,
+      "loss": 0.016,
+      "step": 3870
+    },
+    {
+      "epoch": 0.768894627073195,
+      "grad_norm": 0.4077083813837435,
+      "learning_rate": 1.3360957010419813e-06,
+      "loss": 0.0127,
+      "step": 3871
+    },
+    {
+      "epoch": 0.7690932565299434,
+      "grad_norm": 1.0786672783122584,
+      "learning_rate": 1.333907046743641e-06,
+      "loss": 0.0139,
+      "step": 3872
+    },
+    {
+      "epoch": 0.7692918859866918,
+      "grad_norm": 0.620538079771535,
+      "learning_rate": 1.3317199105707207e-06,
+      "loss": 0.0119,
+      "step": 3873
+    },
+    {
+      "epoch": 0.7694905154434403,
+      "grad_norm": 0.7487056379611181,
+      "learning_rate": 1.3295342934289128e-06,
+      "loss": 0.0135,
+      "step": 3874
+    },
+    {
+      "epoch": 0.7696891449001887,
+      "grad_norm": 0.5658770968139971,
+      "learning_rate": 1.3273501962232787e-06,
+      "loss": 0.0136,
+      "step": 3875
+    },
+    {
+      "epoch": 0.7698877743569371,
+      "grad_norm": 0.5544123537115362,
+      "learning_rate": 1.3251676198582491e-06,
+      "loss": 0.0076,
+      "step": 3876
+    },
+    {
+      "epoch": 0.7700864038136855,
+      "grad_norm": 0.42604725459840853,
+      "learning_rate": 1.3229865652376295e-06,
+      "loss": 0.0092,
+      "step": 3877
+    },
+    {
+      "epoch": 0.770285033270434,
+      "grad_norm": 0.7356682784487021,
+      "learning_rate": 1.3208070332645889e-06,
+      "loss": 0.0091,
+      "step": 3878
+    },
+    {
+      "epoch": 0.7704836627271825,
+      "grad_norm": 0.2945392911865687,
+      "learning_rate": 1.3186290248416723e-06,
+      "loss": 0.0094,
+      "step": 3879
+    },
+    {
+      "epoch": 0.7706822921839309,
+      "grad_norm": 0.4400140082230475,
+      "learning_rate": 1.3164525408707908e-06,
+      "loss": 0.012,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7708809216406793,
+      "grad_norm": 0.3677167541782022,
+      "learning_rate": 1.3142775822532216e-06,
+      "loss": 0.0068,
+      "step": 3881
+    },
+    {
+      "epoch": 0.7710795510974278,
+      "grad_norm": 0.393778558813364,
+      "learning_rate": 1.3121041498896165e-06,
+      "loss": 0.0093,
+      "step": 3882
+    },
+    {
+      "epoch": 0.7712781805541762,
+      "grad_norm": 0.41937295389208945,
+      "learning_rate": 1.3099322446799883e-06,
+      "loss": 0.0109,
+      "step": 3883
+    },
+    {
+      "epoch": 0.7714768100109246,
+      "grad_norm": 0.45418818468109395,
+      "learning_rate": 1.3077618675237235e-06,
+      "loss": 0.009,
+      "step": 3884
+    },
+    {
+      "epoch": 0.771675439467673,
+      "grad_norm": 0.5022438379809215,
+      "learning_rate": 1.305593019319571e-06,
+      "loss": 0.0133,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7718740689244215,
+      "grad_norm": 0.7531062350453083,
+      "learning_rate": 1.3034257009656486e-06,
+      "loss": 0.0098,
+      "step": 3886
+    },
+    {
+      "epoch": 0.7720726983811699,
+      "grad_norm": 0.32685807203315603,
+      "learning_rate": 1.3012599133594438e-06,
+      "loss": 0.0122,
+      "step": 3887
+    },
+    {
+      "epoch": 0.7722713278379184,
+      "grad_norm": 0.6199595695269277,
+      "learning_rate": 1.299095657397803e-06,
+      "loss": 0.0137,
+      "step": 3888
+    },
+    {
+      "epoch": 0.7724699572946668,
+      "grad_norm": 0.918750119255603,
+      "learning_rate": 1.296932933976946e-06,
+      "loss": 0.0183,
+      "step": 3889
+    },
+    {
+      "epoch": 0.7726685867514153,
+      "grad_norm": 0.5007197845896081,
+      "learning_rate": 1.294771743992451e-06,
+      "loss": 0.01,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7728672162081637,
+      "grad_norm": 0.4774983945538157,
+      "learning_rate": 1.292612088339268e-06,
+      "loss": 0.0112,
+      "step": 3891
+    },
+    {
+      "epoch": 0.7730658456649121,
+      "grad_norm": 0.40168632160434964,
+      "learning_rate": 1.2904539679117051e-06,
+      "loss": 0.0045,
+      "step": 3892
+    },
+    {
+      "epoch": 0.7732644751216605,
+      "grad_norm": 0.6613299836977145,
+      "learning_rate": 1.2882973836034391e-06,
+      "loss": 0.0173,
+      "step": 3893
+    },
+    {
+      "epoch": 0.7734631045784089,
+      "grad_norm": 2.4503594172896195,
+      "learning_rate": 1.286142336307511e-06,
+      "loss": 0.016,
+      "step": 3894
+    },
+    {
+      "epoch": 0.7736617340351574,
+      "grad_norm": 0.5175369687816846,
+      "learning_rate": 1.283988826916321e-06,
+      "loss": 0.0129,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7738603634919059,
+      "grad_norm": 0.32059427002808744,
+      "learning_rate": 1.2818368563216377e-06,
+      "loss": 0.0094,
+      "step": 3896
+    },
+    {
+      "epoch": 0.7740589929486543,
+      "grad_norm": 0.45258914822238383,
+      "learning_rate": 1.2796864254145875e-06,
+      "loss": 0.0123,
+      "step": 3897
+    },
+    {
+      "epoch": 0.7742576224054027,
+      "grad_norm": 0.31073316498638376,
+      "learning_rate": 1.277537535085664e-06,
+      "loss": 0.0105,
+      "step": 3898
+    },
+    {
+      "epoch": 0.7744562518621512,
+      "grad_norm": 0.27654739609890827,
+      "learning_rate": 1.2753901862247198e-06,
+      "loss": 0.0068,
+      "step": 3899
+    },
+    {
+      "epoch": 0.7746548813188996,
+      "grad_norm": 0.7702240957803342,
+      "learning_rate": 1.2732443797209676e-06,
+      "loss": 0.014,
+      "step": 3900
+    },
+    {
+      "epoch": 0.774853510775648,
+      "grad_norm": 0.7888401795775241,
+      "learning_rate": 1.2711001164629878e-06,
+      "loss": 0.0107,
+      "step": 3901
+    },
+    {
+      "epoch": 0.7750521402323964,
+      "grad_norm": 0.5102152605501087,
+      "learning_rate": 1.2689573973387136e-06,
+      "loss": 0.0109,
+      "step": 3902
+    },
+    {
+      "epoch": 0.7752507696891449,
+      "grad_norm": 0.42949075002493076,
+      "learning_rate": 1.2668162232354453e-06,
+      "loss": 0.0102,
+      "step": 3903
+    },
+    {
+      "epoch": 0.7754493991458933,
+      "grad_norm": 0.51913928021783,
+      "learning_rate": 1.2646765950398415e-06,
+      "loss": 0.0152,
+      "step": 3904
+    },
+    {
+      "epoch": 0.7756480286026418,
+      "grad_norm": 0.5147647620976568,
+      "learning_rate": 1.2625385136379181e-06,
+      "loss": 0.0109,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7758466580593902,
+      "grad_norm": 0.3620633765074665,
+      "learning_rate": 1.2604019799150547e-06,
+      "loss": 0.0099,
+      "step": 3906
+    },
+    {
+      "epoch": 0.7760452875161387,
+      "grad_norm": 0.623127129178747,
+      "learning_rate": 1.2582669947559845e-06,
+      "loss": 0.0142,
+      "step": 3907
+    },
+    {
+      "epoch": 0.7762439169728871,
+      "grad_norm": 0.8733934868624431,
+      "learning_rate": 1.2561335590448066e-06,
+      "loss": 0.0162,
+      "step": 3908
+    },
+    {
+      "epoch": 0.7764425464296355,
+      "grad_norm": 0.8813493665439702,
+      "learning_rate": 1.2540016736649713e-06,
+      "loss": 0.0126,
+      "step": 3909
+    },
+    {
+      "epoch": 0.7766411758863839,
+      "grad_norm": 0.27843671726361163,
+      "learning_rate": 1.2518713394992916e-06,
+      "loss": 0.0077,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7768398053431324,
+      "grad_norm": 0.5249639405585835,
+      "learning_rate": 1.2497425574299376e-06,
+      "loss": 0.0139,
+      "step": 3911
+    },
+    {
+      "epoch": 0.7770384347998808,
+      "grad_norm": 0.4152029494822117,
+      "learning_rate": 1.247615328338434e-06,
+      "loss": 0.0069,
+      "step": 3912
+    },
+    {
+      "epoch": 0.7772370642566293,
+      "grad_norm": 0.9012103889077504,
+      "learning_rate": 1.2454896531056665e-06,
+      "loss": 0.012,
+      "step": 3913
+    },
+    {
+      "epoch": 0.7774356937133777,
+      "grad_norm": 0.42111435803255814,
+      "learning_rate": 1.2433655326118726e-06,
+      "loss": 0.0102,
+      "step": 3914
+    },
+    {
+      "epoch": 0.7776343231701262,
+      "grad_norm": 0.7641685303436699,
+      "learning_rate": 1.2412429677366512e-06,
+      "loss": 0.0118,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7778329526268746,
+      "grad_norm": 0.5032971538363441,
+      "learning_rate": 1.239121959358951e-06,
+      "loss": 0.0073,
+      "step": 3916
+    },
+    {
+      "epoch": 0.778031582083623,
+      "grad_norm": 0.774857905002289,
+      "learning_rate": 1.2370025083570813e-06,
+      "loss": 0.0082,
+      "step": 3917
+    },
+    {
+      "epoch": 0.7782302115403714,
+      "grad_norm": 0.5279560951716294,
+      "learning_rate": 1.2348846156087058e-06,
+      "loss": 0.0139,
+      "step": 3918
+    },
+    {
+      "epoch": 0.7784288409971198,
+      "grad_norm": 0.4746514702620925,
+      "learning_rate": 1.2327682819908393e-06,
+      "loss": 0.0111,
+      "step": 3919
+    },
+    {
+      "epoch": 0.7786274704538683,
+      "grad_norm": 0.5401006357078507,
+      "learning_rate": 1.230653508379856e-06,
+      "loss": 0.0128,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7788260999106168,
+      "grad_norm": 0.5347132085056566,
+      "learning_rate": 1.2285402956514786e-06,
+      "loss": 0.0094,
+      "step": 3921
+    },
+    {
+      "epoch": 0.7790247293673652,
+      "grad_norm": 0.8006086363779595,
+      "learning_rate": 1.2264286446807893e-06,
+      "loss": 0.0211,
+      "step": 3922
+    },
+    {
+      "epoch": 0.7792233588241136,
+      "grad_norm": 0.729596534109269,
+      "learning_rate": 1.2243185563422194e-06,
+      "loss": 0.0102,
+      "step": 3923
+    },
+    {
+      "epoch": 0.7794219882808621,
+      "grad_norm": 0.8162366400705077,
+      "learning_rate": 1.222210031509553e-06,
+      "loss": 0.0154,
+      "step": 3924
+    },
+    {
+      "epoch": 0.7796206177376105,
+      "grad_norm": 0.8298607900398938,
+      "learning_rate": 1.2201030710559309e-06,
+      "loss": 0.0154,
+      "step": 3925
+    },
+    {
+      "epoch": 0.7798192471943589,
+      "grad_norm": 0.9295004548979635,
+      "learning_rate": 1.2179976758538397e-06,
+      "loss": 0.0202,
+      "step": 3926
+    },
+    {
+      "epoch": 0.7800178766511073,
+      "grad_norm": 0.6284727316819763,
+      "learning_rate": 1.2158938467751258e-06,
+      "loss": 0.0144,
+      "step": 3927
+    },
+    {
+      "epoch": 0.7802165061078558,
+      "grad_norm": 0.31174985864230775,
+      "learning_rate": 1.213791584690978e-06,
+      "loss": 0.008,
+      "step": 3928
+    },
+    {
+      "epoch": 0.7804151355646042,
+      "grad_norm": 0.3904178709197262,
+      "learning_rate": 1.2116908904719443e-06,
+      "loss": 0.008,
+      "step": 3929
+    },
+    {
+      "epoch": 0.7806137650213527,
+      "grad_norm": 0.28385884511776377,
+      "learning_rate": 1.2095917649879202e-06,
+      "loss": 0.0072,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7808123944781011,
+      "grad_norm": 0.8225239608519143,
+      "learning_rate": 1.2074942091081493e-06,
+      "loss": 0.0077,
+      "step": 3931
+    },
+    {
+      "epoch": 0.7810110239348496,
+      "grad_norm": 0.25070849880393475,
+      "learning_rate": 1.2053982237012295e-06,
+      "loss": 0.0067,
+      "step": 3932
+    },
+    {
+      "epoch": 0.781209653391598,
+      "grad_norm": 1.009173744589044,
+      "learning_rate": 1.2033038096351042e-06,
+      "loss": 0.0153,
+      "step": 3933
+    },
+    {
+      "epoch": 0.7814082828483464,
+      "grad_norm": 0.3001079077055634,
+      "learning_rate": 1.20121096777707e-06,
+      "loss": 0.0073,
+      "step": 3934
+    },
+    {
+      "epoch": 0.7816069123050948,
+      "grad_norm": 0.2675585992071396,
+      "learning_rate": 1.1991196989937693e-06,
+      "loss": 0.0077,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7818055417618432,
+      "grad_norm": 0.36211169776646895,
+      "learning_rate": 1.1970300041511945e-06,
+      "loss": 0.0065,
+      "step": 3936
+    },
+    {
+      "epoch": 0.7820041712185917,
+      "grad_norm": 0.556461940995312,
+      "learning_rate": 1.1949418841146875e-06,
+      "loss": 0.0149,
+      "step": 3937
+    },
+    {
+      "epoch": 0.7822028006753402,
+      "grad_norm": 0.41740062523359517,
+      "learning_rate": 1.192855339748935e-06,
+      "loss": 0.0151,
+      "step": 3938
+    },
+    {
+      "epoch": 0.7824014301320886,
+      "grad_norm": 0.31475234587927653,
+      "learning_rate": 1.1907703719179752e-06,
+      "loss": 0.0064,
+      "step": 3939
+    },
+    {
+      "epoch": 0.782600059588837,
+      "grad_norm": 0.6800297738810858,
+      "learning_rate": 1.1886869814851881e-06,
+      "loss": 0.0138,
+      "step": 3940
+    },
+    {
+      "epoch": 0.7827986890455855,
+      "grad_norm": 0.5648602421057914,
+      "learning_rate": 1.186605169313307e-06,
+      "loss": 0.0143,
+      "step": 3941
+    },
+    {
+      "epoch": 0.7829973185023339,
+      "grad_norm": 0.32364785454126055,
+      "learning_rate": 1.1845249362644046e-06,
+      "loss": 0.0111,
+      "step": 3942
+    },
+    {
+      "epoch": 0.7831959479590823,
+      "grad_norm": 7.128471218729785,
+      "learning_rate": 1.1824462831999057e-06,
+      "loss": 0.018,
+      "step": 3943
+    },
+    {
+      "epoch": 0.7833945774158307,
+      "grad_norm": 0.46537352198306225,
+      "learning_rate": 1.1803692109805786e-06,
+      "loss": 0.0096,
+      "step": 3944
+    },
+    {
+      "epoch": 0.7835932068725792,
+      "grad_norm": 0.40237938194467165,
+      "learning_rate": 1.178293720466535e-06,
+      "loss": 0.006,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7837918363293276,
+      "grad_norm": 0.47649481459898346,
+      "learning_rate": 1.1762198125172364e-06,
+      "loss": 0.0135,
+      "step": 3946
+    },
+    {
+      "epoch": 0.7839904657860761,
+      "grad_norm": 0.3889132630569039,
+      "learning_rate": 1.1741474879914837e-06,
+      "loss": 0.0091,
+      "step": 3947
+    },
+    {
+      "epoch": 0.7841890952428245,
+      "grad_norm": 0.3589855854840536,
+      "learning_rate": 1.1720767477474238e-06,
+      "loss": 0.0071,
+      "step": 3948
+    },
+    {
+      "epoch": 0.784387724699573,
+      "grad_norm": 0.6600060042525093,
+      "learning_rate": 1.1700075926425508e-06,
+      "loss": 0.0073,
+      "step": 3949
+    },
+    {
+      "epoch": 0.7845863541563214,
+      "grad_norm": 0.5605141682819981,
+      "learning_rate": 1.167940023533697e-06,
+      "loss": 0.0107,
+      "step": 3950
+    },
+    {
+      "epoch": 0.7847849836130698,
+      "grad_norm": 0.4276017759614575,
+      "learning_rate": 1.1658740412770426e-06,
+      "loss": 0.0108,
+      "step": 3951
+    },
+    {
+      "epoch": 0.7849836130698182,
+      "grad_norm": 0.5245144378224614,
+      "learning_rate": 1.1638096467281074e-06,
+      "loss": 0.0085,
+      "step": 3952
+    },
+    {
+      "epoch": 0.7851822425265667,
+      "grad_norm": 0.504334640539037,
+      "learning_rate": 1.1617468407417553e-06,
+      "loss": 0.0101,
+      "step": 3953
+    },
+    {
+      "epoch": 0.7853808719833151,
+      "grad_norm": 0.48205839823898566,
+      "learning_rate": 1.1596856241721944e-06,
+      "loss": 0.0116,
+      "step": 3954
+    },
+    {
+      "epoch": 0.7855795014400636,
+      "grad_norm": 0.6530170104490158,
+      "learning_rate": 1.1576259978729692e-06,
+      "loss": 0.0178,
+      "step": 3955
+    },
+    {
+      "epoch": 0.785778130896812,
+      "grad_norm": 0.710759087565475,
+      "learning_rate": 1.1555679626969724e-06,
+      "loss": 0.0144,
+      "step": 3956
+    },
+    {
+      "epoch": 0.7859767603535605,
+      "grad_norm": 0.5424274465145213,
+      "learning_rate": 1.1535115194964304e-06,
+      "loss": 0.0132,
+      "step": 3957
+    },
+    {
+      "epoch": 0.7861753898103089,
+      "grad_norm": 0.6661319005283131,
+      "learning_rate": 1.1514566691229178e-06,
+      "loss": 0.013,
+      "step": 3958
+    },
+    {
+      "epoch": 0.7863740192670573,
+      "grad_norm": 0.5573412948277443,
+      "learning_rate": 1.1494034124273428e-06,
+      "loss": 0.0133,
+      "step": 3959
+    },
+    {
+      "epoch": 0.7865726487238057,
+      "grad_norm": 0.4236751414096416,
+      "learning_rate": 1.147351750259959e-06,
+      "loss": 0.0079,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7867712781805541,
+      "grad_norm": 0.6700103105511918,
+      "learning_rate": 1.1453016834703584e-06,
+      "loss": 0.0142,
+      "step": 3961
+    },
+    {
+      "epoch": 0.7869699076373026,
+      "grad_norm": 0.6722312403375201,
+      "learning_rate": 1.1432532129074692e-06,
+      "loss": 0.0175,
+      "step": 3962
+    },
+    {
+      "epoch": 0.7871685370940511,
+      "grad_norm": 0.29540238303248256,
+      "learning_rate": 1.1412063394195634e-06,
+      "loss": 0.0082,
+      "step": 3963
+    },
+    {
+      "epoch": 0.7873671665507995,
+      "grad_norm": 0.453040253508881,
+      "learning_rate": 1.1391610638542473e-06,
+      "loss": 0.0149,
+      "step": 3964
+    },
+    {
+      "epoch": 0.787565796007548,
+      "grad_norm": 0.3063401650275241,
+      "learning_rate": 1.1371173870584696e-06,
+      "loss": 0.0087,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7877644254642964,
+      "grad_norm": 0.7536587487092659,
+      "learning_rate": 1.1350753098785117e-06,
+      "loss": 0.0112,
+      "step": 3966
+    },
+    {
+      "epoch": 0.7879630549210448,
+      "grad_norm": 0.31791285239524447,
+      "learning_rate": 1.1330348331599978e-06,
+      "loss": 0.0101,
+      "step": 3967
+    },
+    {
+      "epoch": 0.7881616843777932,
+      "grad_norm": 0.5345985537470007,
+      "learning_rate": 1.1309959577478885e-06,
+      "loss": 0.0138,
+      "step": 3968
+    },
+    {
+      "epoch": 0.7883603138345416,
+      "grad_norm": 0.39066543567981465,
+      "learning_rate": 1.128958684486477e-06,
+      "loss": 0.0105,
+      "step": 3969
+    },
+    {
+      "epoch": 0.7885589432912901,
+      "grad_norm": 0.6971256838903706,
+      "learning_rate": 1.1269230142194004e-06,
+      "loss": 0.0219,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7887575727480385,
+      "grad_norm": 0.3025242348367761,
+      "learning_rate": 1.1248889477896224e-06,
+      "loss": 0.0064,
+      "step": 3971
+    },
+    {
+      "epoch": 0.788956202204787,
+      "grad_norm": 0.4101701372618817,
+      "learning_rate": 1.1228564860394508e-06,
+      "loss": 0.0085,
+      "step": 3972
+    },
+    {
+      "epoch": 0.7891548316615354,
+      "grad_norm": 1.342044070773984,
+      "learning_rate": 1.120825629810527e-06,
+      "loss": 0.0192,
+      "step": 3973
+    },
+    {
+      "epoch": 0.7893534611182839,
+      "grad_norm": 0.43993593424734334,
+      "learning_rate": 1.1187963799438235e-06,
+      "loss": 0.0124,
+      "step": 3974
+    },
+    {
+      "epoch": 0.7895520905750323,
+      "grad_norm": 0.8408300332951749,
+      "learning_rate": 1.1167687372796537e-06,
+      "loss": 0.0113,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7897507200317807,
+      "grad_norm": 0.9328690964868325,
+      "learning_rate": 1.1147427026576597e-06,
+      "loss": 0.0184,
+      "step": 3976
+    },
+    {
+      "epoch": 0.7899493494885291,
+      "grad_norm": 0.41486356516047107,
+      "learning_rate": 1.1127182769168231e-06,
+      "loss": 0.0048,
+      "step": 3977
+    },
+    {
+      "epoch": 0.7901479789452776,
+      "grad_norm": 0.6855136367785546,
+      "learning_rate": 1.110695460895454e-06,
+      "loss": 0.0129,
+      "step": 3978
+    },
+    {
+      "epoch": 0.790346608402026,
+      "grad_norm": 0.28701281574074733,
+      "learning_rate": 1.108674255431199e-06,
+      "loss": 0.0041,
+      "step": 3979
+    },
+    {
+      "epoch": 0.7905452378587745,
+      "grad_norm": 0.4540954582400495,
+      "learning_rate": 1.10665466136104e-06,
+      "loss": 0.0152,
+      "step": 3980
+    },
+    {
+      "epoch": 0.7907438673155229,
+      "grad_norm": 0.7368225481887282,
+      "learning_rate": 1.1046366795212854e-06,
+      "loss": 0.0125,
+      "step": 3981
+    },
+    {
+      "epoch": 0.7909424967722714,
+      "grad_norm": 0.9802282075745697,
+      "learning_rate": 1.1026203107475824e-06,
+      "loss": 0.0145,
+      "step": 3982
+    },
+    {
+      "epoch": 0.7911411262290198,
+      "grad_norm": 1.0658263094138696,
+      "learning_rate": 1.100605555874904e-06,
+      "loss": 0.0151,
+      "step": 3983
+    },
+    {
+      "epoch": 0.7913397556857682,
+      "grad_norm": 0.35002483862277856,
+      "learning_rate": 1.0985924157375616e-06,
+      "loss": 0.0088,
+      "step": 3984
+    },
+    {
+      "epoch": 0.7915383851425166,
+      "grad_norm": 0.3244457738853698,
+      "learning_rate": 1.0965808911691917e-06,
+      "loss": 0.0066,
+      "step": 3985
+    },
+    {
+      "epoch": 0.791737014599265,
+      "grad_norm": 0.6796365282260834,
+      "learning_rate": 1.0945709830027657e-06,
+      "loss": 0.0094,
+      "step": 3986
+    },
+    {
+      "epoch": 0.7919356440560135,
+      "grad_norm": 0.45854564583815816,
+      "learning_rate": 1.0925626920705857e-06,
+      "loss": 0.0098,
+      "step": 3987
+    },
+    {
+      "epoch": 0.7921342735127619,
+      "grad_norm": 0.5264577666006467,
+      "learning_rate": 1.0905560192042808e-06,
+      "loss": 0.0071,
+      "step": 3988
+    },
+    {
+      "epoch": 0.7923329029695104,
+      "grad_norm": 0.3097781743722543,
+      "learning_rate": 1.0885509652348142e-06,
+      "loss": 0.008,
+      "step": 3989
+    },
+    {
+      "epoch": 0.7925315324262588,
+      "grad_norm": 0.4200986146469633,
+      "learning_rate": 1.086547530992475e-06,
+      "loss": 0.0089,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7927301618830073,
+      "grad_norm": 0.6691123935864665,
+      "learning_rate": 1.0845457173068858e-06,
+      "loss": 0.0151,
+      "step": 3991
+    },
+    {
+      "epoch": 0.7929287913397557,
+      "grad_norm": 0.3305478813634996,
+      "learning_rate": 1.0825455250069921e-06,
+      "loss": 0.005,
+      "step": 3992
+    },
+    {
+      "epoch": 0.7931274207965041,
+      "grad_norm": 0.33748170186551923,
+      "learning_rate": 1.080546954921075e-06,
+      "loss": 0.0062,
+      "step": 3993
+    },
+    {
+      "epoch": 0.7933260502532525,
+      "grad_norm": 0.3245694682203549,
+      "learning_rate": 1.0785500078767392e-06,
+      "loss": 0.0057,
+      "step": 3994
+    },
+    {
+      "epoch": 0.793524679710001,
+      "grad_norm": 0.4644644236708028,
+      "learning_rate": 1.076554684700916e-06,
+      "loss": 0.0089,
+      "step": 3995
+    },
+    {
+      "epoch": 0.7937233091667494,
+      "grad_norm": 0.5038171184311694,
+      "learning_rate": 1.0745609862198692e-06,
+      "loss": 0.0116,
+      "step": 3996
+    },
+    {
+      "epoch": 0.7939219386234979,
+      "grad_norm": 0.6885486523828257,
+      "learning_rate": 1.0725689132591888e-06,
+      "loss": 0.0163,
+      "step": 3997
+    },
+    {
+      "epoch": 0.7941205680802463,
+      "grad_norm": 0.3900499706842763,
+      "learning_rate": 1.0705784666437863e-06,
+      "loss": 0.011,
+      "step": 3998
+    },
+    {
+      "epoch": 0.7943191975369948,
+      "grad_norm": 0.39376689041252166,
+      "learning_rate": 1.0685896471979074e-06,
+      "loss": 0.0072,
+      "step": 3999
+    },
+    {
+      "epoch": 0.7945178269937432,
+      "grad_norm": 0.472464666055705,
+      "learning_rate": 1.066602455745117e-06,
+      "loss": 0.0113,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7947164564504916,
+      "grad_norm": 0.7011028914294587,
+      "learning_rate": 1.0646168931083123e-06,
+      "loss": 0.0159,
+      "step": 4001
+    },
+    {
+      "epoch": 0.79491508590724,
+      "grad_norm": 0.5949751520146886,
+      "learning_rate": 1.0626329601097102e-06,
+      "loss": 0.0188,
+      "step": 4002
+    },
+    {
+      "epoch": 0.7951137153639884,
+      "grad_norm": 0.5403625268434289,
+      "learning_rate": 1.0606506575708559e-06,
+      "loss": 0.0153,
+      "step": 4003
+    },
+    {
+      "epoch": 0.7953123448207369,
+      "grad_norm": 0.32460874730935,
+      "learning_rate": 1.0586699863126205e-06,
+      "loss": 0.0074,
+      "step": 4004
+    },
+    {
+      "epoch": 0.7955109742774854,
+      "grad_norm": 0.5225527629713734,
+      "learning_rate": 1.0566909471551956e-06,
+      "loss": 0.011,
+      "step": 4005
+    },
+    {
+      "epoch": 0.7957096037342338,
+      "grad_norm": 0.23452137300486403,
+      "learning_rate": 1.054713540918102e-06,
+      "loss": 0.006,
+      "step": 4006
+    },
+    {
+      "epoch": 0.7959082331909823,
+      "grad_norm": 0.9970244401372391,
+      "learning_rate": 1.0527377684201788e-06,
+      "loss": 0.0139,
+      "step": 4007
+    },
+    {
+      "epoch": 0.7961068626477307,
+      "grad_norm": 0.5496516753977624,
+      "learning_rate": 1.0507636304795942e-06,
+      "loss": 0.0113,
+      "step": 4008
+    },
+    {
+      "epoch": 0.7963054921044791,
+      "grad_norm": 0.40426703473069714,
+      "learning_rate": 1.0487911279138341e-06,
+      "loss": 0.0115,
+      "step": 4009
+    },
+    {
+      "epoch": 0.7965041215612275,
+      "grad_norm": 0.4432391943593047,
+      "learning_rate": 1.046820261539711e-06,
+      "loss": 0.0087,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7967027510179759,
+      "grad_norm": 0.5434341498818477,
+      "learning_rate": 1.04485103217336e-06,
+      "loss": 0.0133,
+      "step": 4011
+    },
+    {
+      "epoch": 0.7969013804747244,
+      "grad_norm": 0.6869762130705815,
+      "learning_rate": 1.0428834406302345e-06,
+      "loss": 0.0172,
+      "step": 4012
+    },
+    {
+      "epoch": 0.7971000099314728,
+      "grad_norm": 0.3868532857982542,
+      "learning_rate": 1.040917487725115e-06,
+      "loss": 0.0107,
+      "step": 4013
+    },
+    {
+      "epoch": 0.7972986393882213,
+      "grad_norm": 0.7528663077873841,
+      "learning_rate": 1.0389531742720976e-06,
+      "loss": 0.008,
+      "step": 4014
+    },
+    {
+      "epoch": 0.7974972688449697,
+      "grad_norm": 0.47927375494835317,
+      "learning_rate": 1.0369905010846054e-06,
+      "loss": 0.0083,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7976958983017182,
+      "grad_norm": 0.4350659995552571,
+      "learning_rate": 1.0350294689753764e-06,
+      "loss": 0.0088,
+      "step": 4016
+    },
+    {
+      "epoch": 0.7978945277584666,
+      "grad_norm": 0.295174664809995,
+      "learning_rate": 1.0330700787564756e-06,
+      "loss": 0.0045,
+      "step": 4017
+    },
+    {
+      "epoch": 0.798093157215215,
+      "grad_norm": 0.31793807650251443,
+      "learning_rate": 1.0311123312392823e-06,
+      "loss": 0.0099,
+      "step": 4018
+    },
+    {
+      "epoch": 0.7982917866719634,
+      "grad_norm": 0.23482176137295935,
+      "learning_rate": 1.0291562272344968e-06,
+      "loss": 0.0062,
+      "step": 4019
+    },
+    {
+      "epoch": 0.7984904161287119,
+      "grad_norm": 0.5061544706704849,
+      "learning_rate": 1.0272017675521423e-06,
+      "loss": 0.0089,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7986890455854603,
+      "grad_norm": 0.31649122670119934,
+      "learning_rate": 1.0252489530015564e-06,
+      "loss": 0.0064,
+      "step": 4021
+    },
+    {
+      "epoch": 0.7988876750422088,
+      "grad_norm": 0.7586656461806361,
+      "learning_rate": 1.0232977843913983e-06,
+      "loss": 0.0149,
+      "step": 4022
+    },
+    {
+      "epoch": 0.7990863044989572,
+      "grad_norm": 0.6009260380452704,
+      "learning_rate": 1.0213482625296468e-06,
+      "loss": 0.0108,
+      "step": 4023
+    },
+    {
+      "epoch": 0.7992849339557057,
+      "grad_norm": 0.9958776677403033,
+      "learning_rate": 1.0194003882235943e-06,
+      "loss": 0.0113,
+      "step": 4024
+    },
+    {
+      "epoch": 0.7994835634124541,
+      "grad_norm": 0.5106754309922006,
+      "learning_rate": 1.0174541622798556e-06,
+      "loss": 0.0092,
+      "step": 4025
+    },
+    {
+      "epoch": 0.7996821928692025,
+      "grad_norm": 0.5365858874211565,
+      "learning_rate": 1.015509585504359e-06,
+      "loss": 0.0112,
+      "step": 4026
+    },
+    {
+      "epoch": 0.7998808223259509,
+      "grad_norm": 0.5783336315909131,
+      "learning_rate": 1.0135666587023545e-06,
+      "loss": 0.012,
+      "step": 4027
+    },
+    {
+      "epoch": 0.8000794517826993,
+      "grad_norm": 0.7373444717399422,
+      "learning_rate": 1.0116253826784028e-06,
+      "loss": 0.0112,
+      "step": 4028
+    },
+    {
+      "epoch": 0.8002780812394478,
+      "grad_norm": 0.23832355380940837,
+      "learning_rate": 1.0096857582363862e-06,
+      "loss": 0.0037,
+      "step": 4029
+    },
+    {
+      "epoch": 0.8004767106961962,
+      "grad_norm": 0.4162968251704296,
+      "learning_rate": 1.0077477861795026e-06,
+      "loss": 0.0098,
+      "step": 4030
+    },
+    {
+      "epoch": 0.8006753401529447,
+      "grad_norm": 0.2707466456827671,
+      "learning_rate": 1.0058114673102616e-06,
+      "loss": 0.0055,
+      "step": 4031
+    },
+    {
+      "epoch": 0.8008739696096931,
+      "grad_norm": 0.6216931576856428,
+      "learning_rate": 1.0038768024304928e-06,
+      "loss": 0.0139,
+      "step": 4032
+    },
+    {
+      "epoch": 0.8010725990664416,
+      "grad_norm": 0.5832019943429834,
+      "learning_rate": 1.0019437923413373e-06,
+      "loss": 0.0149,
+      "step": 4033
+    },
+    {
+      "epoch": 0.80127122852319,
+      "grad_norm": 0.6152187266424713,
+      "learning_rate": 1.0000124378432553e-06,
+      "loss": 0.0086,
+      "step": 4034
+    },
+    {
+      "epoch": 0.8014698579799384,
+      "grad_norm": 0.3619901915947861,
+      "learning_rate": 9.980827397360155e-07,
+      "loss": 0.0082,
+      "step": 4035
+    },
+    {
+      "epoch": 0.8016684874366868,
+      "grad_norm": 0.3343601589064828,
+      "learning_rate": 9.961546988187055e-07,
+      "loss": 0.0083,
+      "step": 4036
+    },
+    {
+      "epoch": 0.8018671168934353,
+      "grad_norm": 0.5999940967835934,
+      "learning_rate": 9.942283158897264e-07,
+      "loss": 0.0149,
+      "step": 4037
+    },
+    {
+      "epoch": 0.8020657463501837,
+      "grad_norm": 0.7931795315952762,
+      "learning_rate": 9.923035917467887e-07,
+      "loss": 0.0166,
+      "step": 4038
+    },
+    {
+      "epoch": 0.8022643758069322,
+      "grad_norm": 0.3931590426635836,
+      "learning_rate": 9.90380527186922e-07,
+      "loss": 0.0076,
+      "step": 4039
+    },
+    {
+      "epoch": 0.8024630052636806,
+      "grad_norm": 0.4836490308516654,
+      "learning_rate": 9.884591230064622e-07,
+      "loss": 0.0098,
+      "step": 4040
+    },
+    {
+      "epoch": 0.8026616347204291,
+      "grad_norm": 0.473992042830245,
+      "learning_rate": 9.865393800010636e-07,
+      "loss": 0.011,
+      "step": 4041
+    },
+    {
+      "epoch": 0.8028602641771775,
+      "grad_norm": 0.29184782245234137,
+      "learning_rate": 9.84621298965689e-07,
+      "loss": 0.0083,
+      "step": 4042
+    },
+    {
+      "epoch": 0.8030588936339259,
+      "grad_norm": 0.642632073832344,
+      "learning_rate": 9.827048806946115e-07,
+      "loss": 0.0094,
+      "step": 4043
+    },
+    {
+      "epoch": 0.8032575230906743,
+      "grad_norm": 0.3687910450246707,
+      "learning_rate": 9.807901259814211e-07,
+      "loss": 0.0072,
+      "step": 4044
+    },
+    {
+      "epoch": 0.8034561525474228,
+      "grad_norm": 0.37471175625878944,
+      "learning_rate": 9.788770356190137e-07,
+      "loss": 0.0097,
+      "step": 4045
+    },
+    {
+      "epoch": 0.8036547820041712,
+      "grad_norm": 0.357576240436411,
+      "learning_rate": 9.76965610399599e-07,
+      "loss": 0.0065,
+      "step": 4046
+    },
+    {
+      "epoch": 0.8038534114609197,
+      "grad_norm": 0.6214890939496734,
+      "learning_rate": 9.750558511146974e-07,
+      "loss": 0.0069,
+      "step": 4047
+    },
+    {
+      "epoch": 0.8040520409176681,
+      "grad_norm": 1.1538127906781666,
+      "learning_rate": 9.731477585551357e-07,
+      "loss": 0.0152,
+      "step": 4048
+    },
+    {
+      "epoch": 0.8042506703744166,
+      "grad_norm": 0.45686551117881485,
+      "learning_rate": 9.71241333511056e-07,
+      "loss": 0.0082,
+      "step": 4049
+    },
+    {
+      "epoch": 0.804449299831165,
+      "grad_norm": 0.3975400079457581,
+      "learning_rate": 9.693365767719044e-07,
+      "loss": 0.0087,
+      "step": 4050
+    },
+    {
+      "epoch": 0.8046479292879134,
+      "grad_norm": 0.7165138344226095,
+      "learning_rate": 9.674334891264414e-07,
+      "loss": 0.0122,
+      "step": 4051
+    },
+    {
+      "epoch": 0.8048465587446618,
+      "grad_norm": 0.5277203363825931,
+      "learning_rate": 9.65532071362731e-07,
+      "loss": 0.009,
+      "step": 4052
+    },
+    {
+      "epoch": 0.8050451882014102,
+      "grad_norm": 0.6225019351917528,
+      "learning_rate": 9.63632324268149e-07,
+      "loss": 0.0198,
+      "step": 4053
+    },
+    {
+      "epoch": 0.8052438176581587,
+      "grad_norm": 0.4903031415168761,
+      "learning_rate": 9.617342486293812e-07,
+      "loss": 0.0069,
+      "step": 4054
+    },
+    {
+      "epoch": 0.8054424471149071,
+      "grad_norm": 0.5308344689616842,
+      "learning_rate": 9.59837845232416e-07,
+      "loss": 0.0107,
+      "step": 4055
+    },
+    {
+      "epoch": 0.8056410765716556,
+      "grad_norm": 0.7708014862830759,
+      "learning_rate": 9.57943114862554e-07,
+      "loss": 0.0099,
+      "step": 4056
+    },
+    {
+      "epoch": 0.805839706028404,
+      "grad_norm": 1.0172527611136573,
+      "learning_rate": 9.560500583043986e-07,
+      "loss": 0.0087,
+      "step": 4057
+    },
+    {
+      "epoch": 0.8060383354851525,
+      "grad_norm": 0.3705945382686297,
+      "learning_rate": 9.541586763418664e-07,
+      "loss": 0.0122,
+      "step": 4058
+    },
+    {
+      "epoch": 0.8062369649419009,
+      "grad_norm": 0.7158602396036523,
+      "learning_rate": 9.522689697581733e-07,
+      "loss": 0.0117,
+      "step": 4059
+    },
+    {
+      "epoch": 0.8064355943986493,
+      "grad_norm": 0.5429010252706425,
+      "learning_rate": 9.50380939335846e-07,
+      "loss": 0.014,
+      "step": 4060
+    },
+    {
+      "epoch": 0.8066342238553977,
+      "grad_norm": 0.6956019948957465,
+      "learning_rate": 9.48494585856718e-07,
+      "loss": 0.0223,
+      "step": 4061
+    },
+    {
+      "epoch": 0.8068328533121462,
+      "grad_norm": 0.8506079494222205,
+      "learning_rate": 9.466099101019233e-07,
+      "loss": 0.0063,
+      "step": 4062
+    },
+    {
+      "epoch": 0.8070314827688946,
+      "grad_norm": 0.42075834755040786,
+      "learning_rate": 9.447269128519065e-07,
+      "loss": 0.0076,
+      "step": 4063
+    },
+    {
+      "epoch": 0.8072301122256431,
+      "grad_norm": 0.595418453912562,
+      "learning_rate": 9.428455948864134e-07,
+      "loss": 0.0117,
+      "step": 4064
+    },
+    {
+      "epoch": 0.8074287416823915,
+      "grad_norm": 0.6130960282660696,
+      "learning_rate": 9.409659569844975e-07,
+      "loss": 0.0107,
+      "step": 4065
+    },
+    {
+      "epoch": 0.80762737113914,
+      "grad_norm": 1.5118482480538398,
+      "learning_rate": 9.390879999245139e-07,
+      "loss": 0.0109,
+      "step": 4066
+    },
+    {
+      "epoch": 0.8078260005958884,
+      "grad_norm": 0.40907353954507686,
+      "learning_rate": 9.372117244841216e-07,
+      "loss": 0.0077,
+      "step": 4067
+    },
+    {
+      "epoch": 0.8080246300526368,
+      "grad_norm": 0.6408271348047319,
+      "learning_rate": 9.353371314402871e-07,
+      "loss": 0.0112,
+      "step": 4068
+    },
+    {
+      "epoch": 0.8082232595093852,
+      "grad_norm": 0.4828591067512512,
+      "learning_rate": 9.334642215692746e-07,
+      "loss": 0.0095,
+      "step": 4069
+    },
+    {
+      "epoch": 0.8084218889661337,
+      "grad_norm": 0.5431520041134413,
+      "learning_rate": 9.315929956466568e-07,
+      "loss": 0.0139,
+      "step": 4070
+    },
+    {
+      "epoch": 0.8086205184228821,
+      "grad_norm": 0.6310284789435388,
+      "learning_rate": 9.297234544473044e-07,
+      "loss": 0.0113,
+      "step": 4071
+    },
+    {
+      "epoch": 0.8088191478796305,
+      "grad_norm": 0.5081618596537647,
+      "learning_rate": 9.27855598745393e-07,
+      "loss": 0.0099,
+      "step": 4072
+    },
+    {
+      "epoch": 0.809017777336379,
+      "grad_norm": 0.44491804248735184,
+      "learning_rate": 9.259894293144017e-07,
+      "loss": 0.0101,
+      "step": 4073
+    },
+    {
+      "epoch": 0.8092164067931275,
+      "grad_norm": 0.46738266736467293,
+      "learning_rate": 9.241249469271068e-07,
+      "loss": 0.0103,
+      "step": 4074
+    },
+    {
+      "epoch": 0.8094150362498759,
+      "grad_norm": 0.5209162826736811,
+      "learning_rate": 9.222621523555908e-07,
+      "loss": 0.0065,
+      "step": 4075
+    },
+    {
+      "epoch": 0.8096136657066243,
+      "grad_norm": 0.3940989383722829,
+      "learning_rate": 9.204010463712326e-07,
+      "loss": 0.0087,
+      "step": 4076
+    },
+    {
+      "epoch": 0.8098122951633727,
+      "grad_norm": 0.4123877332825086,
+      "learning_rate": 9.18541629744717e-07,
+      "loss": 0.0093,
+      "step": 4077
+    },
+    {
+      "epoch": 0.8100109246201211,
+      "grad_norm": 0.6265895051388984,
+      "learning_rate": 9.16683903246024e-07,
+      "loss": 0.0155,
+      "step": 4078
+    },
+    {
+      "epoch": 0.8102095540768696,
+      "grad_norm": 0.5490533621387226,
+      "learning_rate": 9.148278676444372e-07,
+      "loss": 0.0067,
+      "step": 4079
+    },
+    {
+      "epoch": 0.810408183533618,
+      "grad_norm": 0.3131848581841642,
+      "learning_rate": 9.129735237085408e-07,
+      "loss": 0.0049,
+      "step": 4080
+    },
+    {
+      "epoch": 0.8106068129903665,
+      "grad_norm": 0.7500483326690944,
+      "learning_rate": 9.111208722062143e-07,
+      "loss": 0.0113,
+      "step": 4081
+    },
+    {
+      "epoch": 0.8108054424471149,
+      "grad_norm": 0.5484117682655739,
+      "learning_rate": 9.092699139046413e-07,
+      "loss": 0.011,
+      "step": 4082
+    },
+    {
+      "epoch": 0.8110040719038634,
+      "grad_norm": 0.3098865334339739,
+      "learning_rate": 9.074206495702992e-07,
+      "loss": 0.0086,
+      "step": 4083
+    },
+    {
+      "epoch": 0.8112027013606118,
+      "grad_norm": 0.17468366708042246,
+      "learning_rate": 9.055730799689688e-07,
+      "loss": 0.0028,
+      "step": 4084
+    },
+    {
+      "epoch": 0.8114013308173602,
+      "grad_norm": 0.538338960986114,
+      "learning_rate": 9.037272058657242e-07,
+      "loss": 0.0105,
+      "step": 4085
+    },
+    {
+      "epoch": 0.8115999602741086,
+      "grad_norm": 0.4201409900907841,
+      "learning_rate": 9.018830280249419e-07,
+      "loss": 0.0063,
+      "step": 4086
+    },
+    {
+      "epoch": 0.8117985897308571,
+      "grad_norm": 0.5876635452613788,
+      "learning_rate": 9.000405472102946e-07,
+      "loss": 0.0126,
+      "step": 4087
+    },
+    {
+      "epoch": 0.8119972191876055,
+      "grad_norm": 0.5156736720046141,
+      "learning_rate": 8.981997641847501e-07,
+      "loss": 0.0128,
+      "step": 4088
+    },
+    {
+      "epoch": 0.812195848644354,
+      "grad_norm": 0.5052009076325246,
+      "learning_rate": 8.963606797105767e-07,
+      "loss": 0.015,
+      "step": 4089
+    },
+    {
+      "epoch": 0.8123944781011024,
+      "grad_norm": 0.5430853287081518,
+      "learning_rate": 8.94523294549336e-07,
+      "loss": 0.01,
+      "step": 4090
+    },
+    {
+      "epoch": 0.8125931075578509,
+      "grad_norm": 0.44810523621001547,
+      "learning_rate": 8.92687609461887e-07,
+      "loss": 0.0098,
+      "step": 4091
+    },
+    {
+      "epoch": 0.8127917370145993,
+      "grad_norm": 0.5134814399572584,
+      "learning_rate": 8.908536252083865e-07,
+      "loss": 0.0056,
+      "step": 4092
+    },
+    {
+      "epoch": 0.8129903664713477,
+      "grad_norm": 0.46375774276643456,
+      "learning_rate": 8.890213425482841e-07,
+      "loss": 0.0089,
+      "step": 4093
+    },
+    {
+      "epoch": 0.8131889959280961,
+      "grad_norm": 1.1162249657192767,
+      "learning_rate": 8.871907622403275e-07,
+      "loss": 0.0214,
+      "step": 4094
+    },
+    {
+      "epoch": 0.8133876253848445,
+      "grad_norm": 0.42917047757228455,
+      "learning_rate": 8.853618850425572e-07,
+      "loss": 0.006,
+      "step": 4095
+    },
+    {
+      "epoch": 0.813586254841593,
+      "grad_norm": 0.5208590114708673,
+      "learning_rate": 8.835347117123089e-07,
+      "loss": 0.0161,
+      "step": 4096
+    },
+    {
+      "epoch": 0.8137848842983414,
+      "grad_norm": 0.5308424443047222,
+      "learning_rate": 8.817092430062158e-07,
+      "loss": 0.0161,
+      "step": 4097
+    },
+    {
+      "epoch": 0.8139835137550899,
+      "grad_norm": 0.46028366709106316,
+      "learning_rate": 8.798854796801997e-07,
+      "loss": 0.0097,
+      "step": 4098
+    },
+    {
+      "epoch": 0.8141821432118383,
+      "grad_norm": 0.448761200240888,
+      "learning_rate": 8.780634224894818e-07,
+      "loss": 0.0128,
+      "step": 4099
+    },
+    {
+      "epoch": 0.8143807726685868,
+      "grad_norm": 0.3671637643734095,
+      "learning_rate": 8.762430721885717e-07,
+      "loss": 0.0073,
+      "step": 4100
+    },
+    {
+      "epoch": 0.8145794021253352,
+      "grad_norm": 0.7115867455847144,
+      "learning_rate": 8.74424429531277e-07,
+      "loss": 0.0087,
+      "step": 4101
+    },
+    {
+      "epoch": 0.8147780315820836,
+      "grad_norm": 0.42494740814520693,
+      "learning_rate": 8.726074952706931e-07,
+      "loss": 0.0092,
+      "step": 4102
+    },
+    {
+      "epoch": 0.814976661038832,
+      "grad_norm": 0.7513309884495961,
+      "learning_rate": 8.707922701592126e-07,
+      "loss": 0.008,
+      "step": 4103
+    },
+    {
+      "epoch": 0.8151752904955805,
+      "grad_norm": 1.1383897819360642,
+      "learning_rate": 8.689787549485185e-07,
+      "loss": 0.0138,
+      "step": 4104
+    },
+    {
+      "epoch": 0.8153739199523289,
+      "grad_norm": 0.6557470447168657,
+      "learning_rate": 8.671669503895841e-07,
+      "loss": 0.0198,
+      "step": 4105
+    },
+    {
+      "epoch": 0.8155725494090774,
+      "grad_norm": 0.6581283241574932,
+      "learning_rate": 8.653568572326781e-07,
+      "loss": 0.0148,
+      "step": 4106
+    },
+    {
+      "epoch": 0.8157711788658258,
+      "grad_norm": 0.417139009113113,
+      "learning_rate": 8.635484762273561e-07,
+      "loss": 0.0123,
+      "step": 4107
+    },
+    {
+      "epoch": 0.8159698083225743,
+      "grad_norm": 0.6867991303871931,
+      "learning_rate": 8.617418081224682e-07,
+      "loss": 0.0168,
+      "step": 4108
+    },
+    {
+      "epoch": 0.8161684377793227,
+      "grad_norm": 1.0724616229540893,
+      "learning_rate": 8.599368536661528e-07,
+      "loss": 0.0131,
+      "step": 4109
+    },
+    {
+      "epoch": 0.8163670672360711,
+      "grad_norm": 0.5087955548304802,
+      "learning_rate": 8.581336136058405e-07,
+      "loss": 0.0137,
+      "step": 4110
+    },
+    {
+      "epoch": 0.8165656966928195,
+      "grad_norm": 0.421017074957917,
+      "learning_rate": 8.563320886882514e-07,
+      "loss": 0.0154,
+      "step": 4111
+    },
+    {
+      "epoch": 0.816764326149568,
+      "grad_norm": 0.39496073798864595,
+      "learning_rate": 8.545322796593941e-07,
+      "loss": 0.0157,
+      "step": 4112
+    },
+    {
+      "epoch": 0.8169629556063164,
+      "grad_norm": 0.4997123669546495,
+      "learning_rate": 8.527341872645711e-07,
+      "loss": 0.0143,
+      "step": 4113
+    },
+    {
+      "epoch": 0.8171615850630648,
+      "grad_norm": 0.37415956728088906,
+      "learning_rate": 8.509378122483652e-07,
+      "loss": 0.0102,
+      "step": 4114
+    },
+    {
+      "epoch": 0.8173602145198133,
+      "grad_norm": 0.383596392690376,
+      "learning_rate": 8.491431553546564e-07,
+      "loss": 0.0054,
+      "step": 4115
+    },
+    {
+      "epoch": 0.8175588439765618,
+      "grad_norm": 0.4821601487830514,
+      "learning_rate": 8.473502173266113e-07,
+      "loss": 0.0087,
+      "step": 4116
+    },
+    {
+      "epoch": 0.8177574734333102,
+      "grad_norm": 0.47018873094016517,
+      "learning_rate": 8.455589989066815e-07,
+      "loss": 0.0102,
+      "step": 4117
+    },
+    {
+      "epoch": 0.8179561028900586,
+      "grad_norm": 0.2810129004401643,
+      "learning_rate": 8.437695008366115e-07,
+      "loss": 0.0077,
+      "step": 4118
+    },
+    {
+      "epoch": 0.818154732346807,
+      "grad_norm": 0.5366359137320363,
+      "learning_rate": 8.419817238574273e-07,
+      "loss": 0.0083,
+      "step": 4119
+    },
+    {
+      "epoch": 0.8183533618035554,
+      "grad_norm": 0.316839455464566,
+      "learning_rate": 8.401956687094487e-07,
+      "loss": 0.006,
+      "step": 4120
+    },
+    {
+      "epoch": 0.8185519912603039,
+      "grad_norm": 0.3111926776617216,
+      "learning_rate": 8.384113361322765e-07,
+      "loss": 0.0086,
+      "step": 4121
+    },
+    {
+      "epoch": 0.8187506207170523,
+      "grad_norm": 0.33471856971083414,
+      "learning_rate": 8.366287268648027e-07,
+      "loss": 0.0079,
+      "step": 4122
+    },
+    {
+      "epoch": 0.8189492501738008,
+      "grad_norm": 0.4351278758006974,
+      "learning_rate": 8.348478416452049e-07,
+      "loss": 0.0078,
+      "step": 4123
+    },
+    {
+      "epoch": 0.8191478796305492,
+      "grad_norm": 0.1570818636838516,
+      "learning_rate": 8.330686812109439e-07,
+      "loss": 0.0027,
+      "step": 4124
+    },
+    {
+      "epoch": 0.8193465090872977,
+      "grad_norm": 0.5523549117457559,
+      "learning_rate": 8.312912462987699e-07,
+      "loss": 0.013,
+      "step": 4125
+    },
+    {
+      "epoch": 0.8195451385440461,
+      "grad_norm": 0.3214475082205346,
+      "learning_rate": 8.295155376447151e-07,
+      "loss": 0.0087,
+      "step": 4126
+    },
+    {
+      "epoch": 0.8197437680007945,
+      "grad_norm": 0.29871131133299855,
+      "learning_rate": 8.277415559841012e-07,
+      "loss": 0.0064,
+      "step": 4127
+    },
+    {
+      "epoch": 0.8199423974575429,
+      "grad_norm": 0.9678390162363811,
+      "learning_rate": 8.259693020515292e-07,
+      "loss": 0.0096,
+      "step": 4128
+    },
+    {
+      "epoch": 0.8201410269142914,
+      "grad_norm": 0.460127250128567,
+      "learning_rate": 8.241987765808896e-07,
+      "loss": 0.0128,
+      "step": 4129
+    },
+    {
+      "epoch": 0.8203396563710398,
+      "grad_norm": 0.6057853897482152,
+      "learning_rate": 8.224299803053559e-07,
+      "loss": 0.0088,
+      "step": 4130
+    },
+    {
+      "epoch": 0.8205382858277883,
+      "grad_norm": 0.5206279732652137,
+      "learning_rate": 8.206629139573824e-07,
+      "loss": 0.008,
+      "step": 4131
+    },
+    {
+      "epoch": 0.8207369152845367,
+      "grad_norm": 0.22610796187469387,
+      "learning_rate": 8.188975782687125e-07,
+      "loss": 0.0035,
+      "step": 4132
+    },
+    {
+      "epoch": 0.8209355447412852,
+      "grad_norm": 0.5861117727426646,
+      "learning_rate": 8.171339739703671e-07,
+      "loss": 0.0154,
+      "step": 4133
+    },
+    {
+      "epoch": 0.8211341741980336,
+      "grad_norm": 0.5374366990301481,
+      "learning_rate": 8.153721017926552e-07,
+      "loss": 0.0129,
+      "step": 4134
+    },
+    {
+      "epoch": 0.821332803654782,
+      "grad_norm": 0.4143698403269354,
+      "learning_rate": 8.136119624651645e-07,
+      "loss": 0.0114,
+      "step": 4135
+    },
+    {
+      "epoch": 0.8215314331115304,
+      "grad_norm": 0.37663933807526523,
+      "learning_rate": 8.118535567167673e-07,
+      "loss": 0.0067,
+      "step": 4136
+    },
+    {
+      "epoch": 0.8217300625682789,
+      "grad_norm": 0.9518297277488379,
+      "learning_rate": 8.100968852756208e-07,
+      "loss": 0.0155,
+      "step": 4137
+    },
+    {
+      "epoch": 0.8219286920250273,
+      "grad_norm": 0.4973523714635241,
+      "learning_rate": 8.083419488691563e-07,
+      "loss": 0.0108,
+      "step": 4138
+    },
+    {
+      "epoch": 0.8221273214817757,
+      "grad_norm": 0.4128944945882541,
+      "learning_rate": 8.065887482240925e-07,
+      "loss": 0.0105,
+      "step": 4139
+    },
+    {
+      "epoch": 0.8223259509385242,
+      "grad_norm": 0.43279127906825904,
+      "learning_rate": 8.048372840664298e-07,
+      "loss": 0.0086,
+      "step": 4140
+    },
+    {
+      "epoch": 0.8225245803952727,
+      "grad_norm": 0.5404311137276034,
+      "learning_rate": 8.030875571214458e-07,
+      "loss": 0.0118,
+      "step": 4141
+    },
+    {
+      "epoch": 0.8227232098520211,
+      "grad_norm": 0.6347070223716965,
+      "learning_rate": 8.013395681137027e-07,
+      "loss": 0.0154,
+      "step": 4142
+    },
+    {
+      "epoch": 0.8229218393087695,
+      "grad_norm": 0.44237408691615476,
+      "learning_rate": 7.995933177670385e-07,
+      "loss": 0.0108,
+      "step": 4143
+    },
+    {
+      "epoch": 0.8231204687655179,
+      "grad_norm": 0.6010470786401767,
+      "learning_rate": 7.978488068045764e-07,
+      "loss": 0.015,
+      "step": 4144
+    },
+    {
+      "epoch": 0.8233190982222663,
+      "grad_norm": 0.7701565584935676,
+      "learning_rate": 7.961060359487138e-07,
+      "loss": 0.0119,
+      "step": 4145
+    },
+    {
+      "epoch": 0.8235177276790148,
+      "grad_norm": 0.4310605835446213,
+      "learning_rate": 7.943650059211322e-07,
+      "loss": 0.0081,
+      "step": 4146
+    },
+    {
+      "epoch": 0.8237163571357632,
+      "grad_norm": 0.590740099431193,
+      "learning_rate": 7.92625717442791e-07,
+      "loss": 0.0167,
+      "step": 4147
+    },
+    {
+      "epoch": 0.8239149865925117,
+      "grad_norm": 0.9726708073139096,
+      "learning_rate": 7.908881712339256e-07,
+      "loss": 0.0189,
+      "step": 4148
+    },
+    {
+      "epoch": 0.8241136160492601,
+      "grad_norm": 0.4053559892400188,
+      "learning_rate": 7.891523680140545e-07,
+      "loss": 0.007,
+      "step": 4149
+    },
+    {
+      "epoch": 0.8243122455060086,
+      "grad_norm": 0.2627943863024738,
+      "learning_rate": 7.874183085019698e-07,
+      "loss": 0.0075,
+      "step": 4150
+    },
+    {
+      "epoch": 0.824510874962757,
+      "grad_norm": 0.6733876425443319,
+      "learning_rate": 7.856859934157463e-07,
+      "loss": 0.0113,
+      "step": 4151
+    },
+    {
+      "epoch": 0.8247095044195054,
+      "grad_norm": 0.42436341922801024,
+      "learning_rate": 7.839554234727309e-07,
+      "loss": 0.0105,
+      "step": 4152
+    },
+    {
+      "epoch": 0.8249081338762538,
+      "grad_norm": 0.6937300129824238,
+      "learning_rate": 7.822265993895533e-07,
+      "loss": 0.01,
+      "step": 4153
+    },
+    {
+      "epoch": 0.8251067633330023,
+      "grad_norm": 0.3959765854010693,
+      "learning_rate": 7.804995218821182e-07,
+      "loss": 0.0069,
+      "step": 4154
+    },
+    {
+      "epoch": 0.8253053927897507,
+      "grad_norm": 0.5846032872762472,
+      "learning_rate": 7.787741916656038e-07,
+      "loss": 0.0091,
+      "step": 4155
+    },
+    {
+      "epoch": 0.8255040222464991,
+      "grad_norm": 0.6565011551242094,
+      "learning_rate": 7.770506094544711e-07,
+      "loss": 0.0107,
+      "step": 4156
+    },
+    {
+      "epoch": 0.8257026517032476,
+      "grad_norm": 0.612137058681032,
+      "learning_rate": 7.753287759624506e-07,
+      "loss": 0.0123,
+      "step": 4157
+    },
+    {
+      "epoch": 0.8259012811599961,
+      "grad_norm": 0.5434012901407049,
+      "learning_rate": 7.736086919025549e-07,
+      "loss": 0.0098,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8260999106167445,
+      "grad_norm": 0.45931089115946994,
+      "learning_rate": 7.718903579870656e-07,
+      "loss": 0.0132,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8262985400734929,
+      "grad_norm": 0.3361927264054978,
+      "learning_rate": 7.701737749275457e-07,
+      "loss": 0.0074,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8264971695302413,
+      "grad_norm": 0.5681567318696761,
+      "learning_rate": 7.684589434348316e-07,
+      "loss": 0.0092,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8266957989869897,
+      "grad_norm": 0.5567095462260514,
+      "learning_rate": 7.66745864219029e-07,
+      "loss": 0.0152,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8268944284437382,
+      "grad_norm": 0.6126892428547912,
+      "learning_rate": 7.650345379895263e-07,
+      "loss": 0.0118,
+      "step": 4163
+    },
+    {
+      "epoch": 0.8270930579004866,
+      "grad_norm": 0.8912645228011347,
+      "learning_rate": 7.63324965454979e-07,
+      "loss": 0.0145,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8272916873572351,
+      "grad_norm": 0.5553683620343327,
+      "learning_rate": 7.616171473233208e-07,
+      "loss": 0.0127,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8274903168139836,
+      "grad_norm": 0.5464361540926801,
+      "learning_rate": 7.599110843017588e-07,
+      "loss": 0.0196,
+      "step": 4166
+    },
+    {
+      "epoch": 0.827688946270732,
+      "grad_norm": 0.7188318229017358,
+      "learning_rate": 7.582067770967694e-07,
+      "loss": 0.0091,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8278875757274804,
+      "grad_norm": 0.6648555543168041,
+      "learning_rate": 7.565042264141071e-07,
+      "loss": 0.0161,
+      "step": 4168
+    },
+    {
+      "epoch": 0.8280862051842288,
+      "grad_norm": 0.5054813315229681,
+      "learning_rate": 7.548034329587934e-07,
+      "loss": 0.0143,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8282848346409772,
+      "grad_norm": 0.3704689407540544,
+      "learning_rate": 7.531043974351282e-07,
+      "loss": 0.0096,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8284834640977257,
+      "grad_norm": 0.5217578074725935,
+      "learning_rate": 7.514071205466783e-07,
+      "loss": 0.0092,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8286820935544741,
+      "grad_norm": 0.4232794133302012,
+      "learning_rate": 7.497116029962848e-07,
+      "loss": 0.0113,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8288807230112225,
+      "grad_norm": 0.5154993721621637,
+      "learning_rate": 7.480178454860615e-07,
+      "loss": 0.0137,
+      "step": 4173
+    },
+    {
+      "epoch": 0.829079352467971,
+      "grad_norm": 0.7002762553322276,
+      "learning_rate": 7.463258487173891e-07,
+      "loss": 0.0122,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8292779819247195,
+      "grad_norm": 0.2849167502483706,
+      "learning_rate": 7.446356133909244e-07,
+      "loss": 0.0053,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8294766113814679,
+      "grad_norm": 0.3481357192610811,
+      "learning_rate": 7.429471402065891e-07,
+      "loss": 0.0057,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8296752408382163,
+      "grad_norm": 0.5261029513879698,
+      "learning_rate": 7.412604298635817e-07,
+      "loss": 0.0142,
+      "step": 4177
+    },
+    {
+      "epoch": 0.8298738702949647,
+      "grad_norm": 0.43848638143307833,
+      "learning_rate": 7.395754830603636e-07,
+      "loss": 0.0083,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8300724997517132,
+      "grad_norm": 0.47296015769234956,
+      "learning_rate": 7.37892300494672e-07,
+      "loss": 0.0129,
+      "step": 4179
+    },
+    {
+      "epoch": 0.8302711292084616,
+      "grad_norm": 0.2953837424429043,
+      "learning_rate": 7.362108828635117e-07,
+      "loss": 0.0095,
+      "step": 4180
+    },
+    {
+      "epoch": 0.83046975866521,
+      "grad_norm": 0.3474888761012372,
+      "learning_rate": 7.345312308631536e-07,
+      "loss": 0.0084,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8306683881219585,
+      "grad_norm": 0.4370889556934582,
+      "learning_rate": 7.328533451891423e-07,
+      "loss": 0.011,
+      "step": 4182
+    },
+    {
+      "epoch": 0.830867017578707,
+      "grad_norm": 0.4255737427042872,
+      "learning_rate": 7.311772265362866e-07,
+      "loss": 0.0133,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8310656470354554,
+      "grad_norm": 0.3946025875468651,
+      "learning_rate": 7.295028755986678e-07,
+      "loss": 0.0093,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8312642764922038,
+      "grad_norm": 0.40631146235195387,
+      "learning_rate": 7.278302930696312e-07,
+      "loss": 0.0105,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8314629059489522,
+      "grad_norm": 0.37683409726127903,
+      "learning_rate": 7.261594796417915e-07,
+      "loss": 0.016,
+      "step": 4186
+    },
+    {
+      "epoch": 0.8316615354057006,
+      "grad_norm": 0.5759629439702126,
+      "learning_rate": 7.244904360070321e-07,
+      "loss": 0.0093,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8318601648624491,
+      "grad_norm": 0.4095970190194349,
+      "learning_rate": 7.228231628565003e-07,
+      "loss": 0.0089,
+      "step": 4188
+    },
+    {
+      "epoch": 0.8320587943191975,
+      "grad_norm": 0.8007934056507748,
+      "learning_rate": 7.211576608806132e-07,
+      "loss": 0.0129,
+      "step": 4189
+    },
+    {
+      "epoch": 0.832257423775946,
+      "grad_norm": 0.4194690457518919,
+      "learning_rate": 7.194939307690557e-07,
+      "loss": 0.0088,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8324560532326944,
+      "grad_norm": 0.4443434778514458,
+      "learning_rate": 7.17831973210773e-07,
+      "loss": 0.0093,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8326546826894429,
+      "grad_norm": 0.3928978159389848,
+      "learning_rate": 7.161717888939834e-07,
+      "loss": 0.0095,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8328533121461913,
+      "grad_norm": 0.431478336303014,
+      "learning_rate": 7.145133785061648e-07,
+      "loss": 0.0105,
+      "step": 4193
+    },
+    {
+      "epoch": 0.8330519416029397,
+      "grad_norm": 0.5225194392334359,
+      "learning_rate": 7.12856742734066e-07,
+      "loss": 0.0196,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8332505710596881,
+      "grad_norm": 0.36971643133265614,
+      "learning_rate": 7.112018822636951e-07,
+      "loss": 0.0079,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8334492005164366,
+      "grad_norm": 0.6234502273500208,
+      "learning_rate": 7.095487977803306e-07,
+      "loss": 0.0126,
+      "step": 4196
+    },
+    {
+      "epoch": 0.833647829973185,
+      "grad_norm": 0.4720027613289749,
+      "learning_rate": 7.078974899685132e-07,
+      "loss": 0.0145,
+      "step": 4197
+    },
+    {
+      "epoch": 0.8338464594299334,
+      "grad_norm": 0.42761253917678405,
+      "learning_rate": 7.062479595120458e-07,
+      "loss": 0.0095,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8340450888866819,
+      "grad_norm": 0.2903106453517186,
+      "learning_rate": 7.046002070939995e-07,
+      "loss": 0.0076,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8342437183434304,
+      "grad_norm": 0.5331366190524331,
+      "learning_rate": 7.029542333967049e-07,
+      "loss": 0.0138,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8344423478001788,
+      "grad_norm": 0.3941782778497735,
+      "learning_rate": 7.013100391017602e-07,
+      "loss": 0.0053,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8346409772569272,
+      "grad_norm": 0.5155012599849433,
+      "learning_rate": 6.996676248900219e-07,
+      "loss": 0.0119,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8348396067136756,
+      "grad_norm": 0.6437821506076101,
+      "learning_rate": 6.980269914416144e-07,
+      "loss": 0.0117,
+      "step": 4203
+    },
+    {
+      "epoch": 0.835038236170424,
+      "grad_norm": 0.24442114698882744,
+      "learning_rate": 6.963881394359223e-07,
+      "loss": 0.004,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8352368656271725,
+      "grad_norm": 0.9459746769962994,
+      "learning_rate": 6.947510695515913e-07,
+      "loss": 0.0263,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8354354950839209,
+      "grad_norm": 0.2833211575993794,
+      "learning_rate": 6.931157824665319e-07,
+      "loss": 0.0066,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8356341245406694,
+      "grad_norm": 0.4356603803658812,
+      "learning_rate": 6.914822788579123e-07,
+      "loss": 0.01,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8358327539974179,
+      "grad_norm": 0.5338876349294479,
+      "learning_rate": 6.898505594021681e-07,
+      "loss": 0.0127,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8360313834541663,
+      "grad_norm": 0.5494695551352939,
+      "learning_rate": 6.882206247749907e-07,
+      "loss": 0.012,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8362300129109147,
+      "grad_norm": 0.6861938545700319,
+      "learning_rate": 6.865924756513336e-07,
+      "loss": 0.0191,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8364286423676631,
+      "grad_norm": 0.4886731636680826,
+      "learning_rate": 6.849661127054141e-07,
+      "loss": 0.009,
+      "step": 4211
+    },
+    {
+      "epoch": 0.8366272718244115,
+      "grad_norm": 0.3818112246211818,
+      "learning_rate": 6.833415366107049e-07,
+      "loss": 0.0103,
+      "step": 4212
+    },
+    {
+      "epoch": 0.83682590128116,
+      "grad_norm": 0.21974960483824155,
+      "learning_rate": 6.817187480399434e-07,
+      "loss": 0.0045,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8370245307379084,
+      "grad_norm": 0.46671217892639616,
+      "learning_rate": 6.800977476651232e-07,
+      "loss": 0.0128,
+      "step": 4214
+    },
+    {
+      "epoch": 0.8372231601946568,
+      "grad_norm": 0.6528384272178425,
+      "learning_rate": 6.784785361574997e-07,
+      "loss": 0.0114,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8374217896514053,
+      "grad_norm": 0.4510583456447262,
+      "learning_rate": 6.768611141875875e-07,
+      "loss": 0.0088,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8376204191081538,
+      "grad_norm": 0.40374865267390264,
+      "learning_rate": 6.752454824251575e-07,
+      "loss": 0.011,
+      "step": 4217
+    },
+    {
+      "epoch": 0.8378190485649022,
+      "grad_norm": 0.41459767009959575,
+      "learning_rate": 6.736316415392435e-07,
+      "loss": 0.0093,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8380176780216506,
+      "grad_norm": 0.2681325117645612,
+      "learning_rate": 6.720195921981332e-07,
+      "loss": 0.0038,
+      "step": 4219
+    },
+    {
+      "epoch": 0.838216307478399,
+      "grad_norm": 0.42508478595734384,
+      "learning_rate": 6.704093350693763e-07,
+      "loss": 0.015,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8384149369351475,
+      "grad_norm": 0.37595114785371714,
+      "learning_rate": 6.688008708197774e-07,
+      "loss": 0.0073,
+      "step": 4221
+    },
+    {
+      "epoch": 0.8386135663918959,
+      "grad_norm": 0.31877225252641367,
+      "learning_rate": 6.671942001154003e-07,
+      "loss": 0.0067,
+      "step": 4222
+    },
+    {
+      "epoch": 0.8388121958486443,
+      "grad_norm": 0.46674482672364,
+      "learning_rate": 6.655893236215671e-07,
+      "loss": 0.0135,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8390108253053928,
+      "grad_norm": 0.41001893376938964,
+      "learning_rate": 6.639862420028531e-07,
+      "loss": 0.0105,
+      "step": 4224
+    },
+    {
+      "epoch": 0.8392094547621413,
+      "grad_norm": 0.7742237052495848,
+      "learning_rate": 6.623849559230955e-07,
+      "loss": 0.0171,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8394080842188897,
+      "grad_norm": 1.0107664886273078,
+      "learning_rate": 6.607854660453827e-07,
+      "loss": 0.0148,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8396067136756381,
+      "grad_norm": 0.7350900223607449,
+      "learning_rate": 6.591877730320639e-07,
+      "loss": 0.0105,
+      "step": 4227
+    },
+    {
+      "epoch": 0.8398053431323865,
+      "grad_norm": 0.6237840242821349,
+      "learning_rate": 6.575918775447404e-07,
+      "loss": 0.0104,
+      "step": 4228
+    },
+    {
+      "epoch": 0.840003972589135,
+      "grad_norm": 0.2205986767974159,
+      "learning_rate": 6.559977802442719e-07,
+      "loss": 0.0079,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8402026020458834,
+      "grad_norm": 0.391922815119898,
+      "learning_rate": 6.54405481790773e-07,
+      "loss": 0.0071,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8404012315026318,
+      "grad_norm": 0.6980785689595316,
+      "learning_rate": 6.52814982843612e-07,
+      "loss": 0.0181,
+      "step": 4231
+    },
+    {
+      "epoch": 0.8405998609593803,
+      "grad_norm": 0.8308250329575882,
+      "learning_rate": 6.512262840614137e-07,
+      "loss": 0.0107,
+      "step": 4232
+    },
+    {
+      "epoch": 0.8407984904161288,
+      "grad_norm": 0.5525708162928655,
+      "learning_rate": 6.496393861020562e-07,
+      "loss": 0.0099,
+      "step": 4233
+    },
+    {
+      "epoch": 0.8409971198728772,
+      "grad_norm": 0.4976875182817474,
+      "learning_rate": 6.480542896226716e-07,
+      "loss": 0.0083,
+      "step": 4234
+    },
+    {
+      "epoch": 0.8411957493296256,
+      "grad_norm": 0.34387989751239384,
+      "learning_rate": 6.464709952796482e-07,
+      "loss": 0.0082,
+      "step": 4235
+    },
+    {
+      "epoch": 0.841394378786374,
+      "grad_norm": 0.5003515193396486,
+      "learning_rate": 6.448895037286251e-07,
+      "loss": 0.0079,
+      "step": 4236
+    },
+    {
+      "epoch": 0.8415930082431224,
+      "grad_norm": 0.476787360201741,
+      "learning_rate": 6.433098156244982e-07,
+      "loss": 0.0098,
+      "step": 4237
+    },
+    {
+      "epoch": 0.8417916376998709,
+      "grad_norm": 0.5623162439430519,
+      "learning_rate": 6.417319316214126e-07,
+      "loss": 0.0124,
+      "step": 4238
+    },
+    {
+      "epoch": 0.8419902671566193,
+      "grad_norm": 0.368015208338776,
+      "learning_rate": 6.401558523727703e-07,
+      "loss": 0.0098,
+      "step": 4239
+    },
+    {
+      "epoch": 0.8421888966133677,
+      "grad_norm": 0.5533902338765675,
+      "learning_rate": 6.385815785312238e-07,
+      "loss": 0.012,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8423875260701162,
+      "grad_norm": 0.4726896319372486,
+      "learning_rate": 6.37009110748677e-07,
+      "loss": 0.0139,
+      "step": 4241
+    },
+    {
+      "epoch": 0.8425861555268647,
+      "grad_norm": 0.4250515882151428,
+      "learning_rate": 6.354384496762894e-07,
+      "loss": 0.01,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8427847849836131,
+      "grad_norm": 0.25361240255528045,
+      "learning_rate": 6.33869595964468e-07,
+      "loss": 0.0069,
+      "step": 4243
+    },
+    {
+      "epoch": 0.8429834144403615,
+      "grad_norm": 0.44441454060510527,
+      "learning_rate": 6.323025502628754e-07,
+      "loss": 0.0095,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8431820438971099,
+      "grad_norm": 0.599605477417074,
+      "learning_rate": 6.30737313220422e-07,
+      "loss": 0.0091,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8433806733538584,
+      "grad_norm": 0.4374505279129274,
+      "learning_rate": 6.291738854852719e-07,
+      "loss": 0.0079,
+      "step": 4246
+    },
+    {
+      "epoch": 0.8435793028106068,
+      "grad_norm": 0.7710884713360233,
+      "learning_rate": 6.276122677048396e-07,
+      "loss": 0.0135,
+      "step": 4247
+    },
+    {
+      "epoch": 0.8437779322673552,
+      "grad_norm": 0.5753325234304452,
+      "learning_rate": 6.260524605257873e-07,
+      "loss": 0.0148,
+      "step": 4248
+    },
+    {
+      "epoch": 0.8439765617241037,
+      "grad_norm": 0.5226453715532203,
+      "learning_rate": 6.244944645940326e-07,
+      "loss": 0.013,
+      "step": 4249
+    },
+    {
+      "epoch": 0.8441751911808522,
+      "grad_norm": 0.4733145956448767,
+      "learning_rate": 6.229382805547369e-07,
+      "loss": 0.0142,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8443738206376006,
+      "grad_norm": 0.7137863419541726,
+      "learning_rate": 6.213839090523166e-07,
+      "loss": 0.0124,
+      "step": 4251
+    },
+    {
+      "epoch": 0.844572450094349,
+      "grad_norm": 0.5052172387349066,
+      "learning_rate": 6.198313507304343e-07,
+      "loss": 0.0214,
+      "step": 4252
+    },
+    {
+      "epoch": 0.8447710795510974,
+      "grad_norm": 0.36751253905346604,
+      "learning_rate": 6.182806062320029e-07,
+      "loss": 0.0086,
+      "step": 4253
+    },
+    {
+      "epoch": 0.8449697090078458,
+      "grad_norm": 0.6136397561793601,
+      "learning_rate": 6.167316761991854e-07,
+      "loss": 0.0144,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8451683384645943,
+      "grad_norm": 1.413194361081298,
+      "learning_rate": 6.151845612733909e-07,
+      "loss": 0.0066,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8453669679213427,
+      "grad_norm": 0.38786949863789816,
+      "learning_rate": 6.136392620952791e-07,
+      "loss": 0.0148,
+      "step": 4256
+    },
+    {
+      "epoch": 0.8455655973780911,
+      "grad_norm": 0.3696999820971262,
+      "learning_rate": 6.120957793047561e-07,
+      "loss": 0.0065,
+      "step": 4257
+    },
+    {
+      "epoch": 0.8457642268348396,
+      "grad_norm": 0.5675751446068034,
+      "learning_rate": 6.105541135409759e-07,
+      "loss": 0.0134,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8459628562915881,
+      "grad_norm": 0.49713563656670917,
+      "learning_rate": 6.090142654423425e-07,
+      "loss": 0.0076,
+      "step": 4259
+    },
+    {
+      "epoch": 0.8461614857483365,
+      "grad_norm": 0.36177582839219674,
+      "learning_rate": 6.074762356465036e-07,
+      "loss": 0.0101,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8463601152050849,
+      "grad_norm": 0.682481331309334,
+      "learning_rate": 6.05940024790358e-07,
+      "loss": 0.0127,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8465587446618333,
+      "grad_norm": 0.3292789124528157,
+      "learning_rate": 6.044056335100462e-07,
+      "loss": 0.006,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8467573741185818,
+      "grad_norm": 0.64966583910457,
+      "learning_rate": 6.028730624409612e-07,
+      "loss": 0.0116,
+      "step": 4263
+    },
+    {
+      "epoch": 0.8469560035753302,
+      "grad_norm": 0.8237702138641505,
+      "learning_rate": 6.013423122177364e-07,
+      "loss": 0.0103,
+      "step": 4264
+    },
+    {
+      "epoch": 0.8471546330320786,
+      "grad_norm": 0.39478232550170367,
+      "learning_rate": 5.998133834742553e-07,
+      "loss": 0.0058,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8473532624888271,
+      "grad_norm": 0.6707468045773747,
+      "learning_rate": 5.982862768436464e-07,
+      "loss": 0.0093,
+      "step": 4266
+    },
+    {
+      "epoch": 0.8475518919455756,
+      "grad_norm": 0.7479799316317778,
+      "learning_rate": 5.967609929582818e-07,
+      "loss": 0.0226,
+      "step": 4267
+    },
+    {
+      "epoch": 0.847750521402324,
+      "grad_norm": 0.6860784103609919,
+      "learning_rate": 5.952375324497811e-07,
+      "loss": 0.0106,
+      "step": 4268
+    },
+    {
+      "epoch": 0.8479491508590724,
+      "grad_norm": 0.5184502971263298,
+      "learning_rate": 5.937158959490064e-07,
+      "loss": 0.0086,
+      "step": 4269
+    },
+    {
+      "epoch": 0.8481477803158208,
+      "grad_norm": 0.39623222920861734,
+      "learning_rate": 5.921960840860674e-07,
+      "loss": 0.0142,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8483464097725693,
+      "grad_norm": 2.3172804734227426,
+      "learning_rate": 5.906780974903154e-07,
+      "loss": 0.0167,
+      "step": 4271
+    },
+    {
+      "epoch": 0.8485450392293177,
+      "grad_norm": 0.3739822933902876,
+      "learning_rate": 5.891619367903467e-07,
+      "loss": 0.0104,
+      "step": 4272
+    },
+    {
+      "epoch": 0.8487436686860661,
+      "grad_norm": 0.3235758648301256,
+      "learning_rate": 5.876476026140043e-07,
+      "loss": 0.0065,
+      "step": 4273
+    },
+    {
+      "epoch": 0.8489422981428146,
+      "grad_norm": 0.4318882083611534,
+      "learning_rate": 5.861350955883705e-07,
+      "loss": 0.011,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8491409275995631,
+      "grad_norm": 1.0955982519259333,
+      "learning_rate": 5.846244163397741e-07,
+      "loss": 0.0139,
+      "step": 4275
+    },
+    {
+      "epoch": 0.8493395570563115,
+      "grad_norm": 0.5712222512112204,
+      "learning_rate": 5.831155654937837e-07,
+      "loss": 0.0149,
+      "step": 4276
+    },
+    {
+      "epoch": 0.8495381865130599,
+      "grad_norm": 0.3652696761055196,
+      "learning_rate": 5.816085436752162e-07,
+      "loss": 0.0064,
+      "step": 4277
+    },
+    {
+      "epoch": 0.8497368159698083,
+      "grad_norm": 0.7739356696768167,
+      "learning_rate": 5.80103351508125e-07,
+      "loss": 0.0139,
+      "step": 4278
+    },
+    {
+      "epoch": 0.8499354454265567,
+      "grad_norm": 0.42583485037192603,
+      "learning_rate": 5.785999896158096e-07,
+      "loss": 0.0092,
+      "step": 4279
+    },
+    {
+      "epoch": 0.8501340748833052,
+      "grad_norm": 0.6344045019567613,
+      "learning_rate": 5.770984586208112e-07,
+      "loss": 0.0082,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8503327043400536,
+      "grad_norm": 0.6040578317404494,
+      "learning_rate": 5.755987591449125e-07,
+      "loss": 0.0127,
+      "step": 4281
+    },
+    {
+      "epoch": 0.850531333796802,
+      "grad_norm": 0.4502937353117837,
+      "learning_rate": 5.741008918091362e-07,
+      "loss": 0.0113,
+      "step": 4282
+    },
+    {
+      "epoch": 0.8507299632535505,
+      "grad_norm": 0.4734975385054619,
+      "learning_rate": 5.726048572337489e-07,
+      "loss": 0.0123,
+      "step": 4283
+    },
+    {
+      "epoch": 0.850928592710299,
+      "grad_norm": 0.335158237315028,
+      "learning_rate": 5.711106560382556e-07,
+      "loss": 0.0071,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8511272221670474,
+      "grad_norm": 0.6315586556303954,
+      "learning_rate": 5.696182888414053e-07,
+      "loss": 0.0202,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8513258516237958,
+      "grad_norm": 0.5356384799146305,
+      "learning_rate": 5.681277562611842e-07,
+      "loss": 0.013,
+      "step": 4286
+    },
+    {
+      "epoch": 0.8515244810805442,
+      "grad_norm": 0.5933153411083376,
+      "learning_rate": 5.666390589148219e-07,
+      "loss": 0.0101,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8517231105372927,
+      "grad_norm": 0.29937152219167246,
+      "learning_rate": 5.651521974187846e-07,
+      "loss": 0.0083,
+      "step": 4288
+    },
+    {
+      "epoch": 0.8519217399940411,
+      "grad_norm": 0.7624990790569742,
+      "learning_rate": 5.636671723887816e-07,
+      "loss": 0.0132,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8521203694507895,
+      "grad_norm": 0.18558021382686216,
+      "learning_rate": 5.621839844397603e-07,
+      "loss": 0.0047,
+      "step": 4290
+    },
+    {
+      "epoch": 0.852318998907538,
+      "grad_norm": 0.31218589825449133,
+      "learning_rate": 5.607026341859062e-07,
+      "loss": 0.0094,
+      "step": 4291
+    },
+    {
+      "epoch": 0.8525176283642865,
+      "grad_norm": 0.3659669083845685,
+      "learning_rate": 5.592231222406463e-07,
+      "loss": 0.0033,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8527162578210349,
+      "grad_norm": 0.5248556606937119,
+      "learning_rate": 5.57745449216644e-07,
+      "loss": 0.0075,
+      "step": 4293
+    },
+    {
+      "epoch": 0.8529148872777833,
+      "grad_norm": 0.30725668244669896,
+      "learning_rate": 5.562696157258029e-07,
+      "loss": 0.0075,
+      "step": 4294
+    },
+    {
+      "epoch": 0.8531135167345317,
+      "grad_norm": 0.5613863285877204,
+      "learning_rate": 5.547956223792633e-07,
+      "loss": 0.0081,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8533121461912802,
+      "grad_norm": 0.2473969543840592,
+      "learning_rate": 5.533234697874045e-07,
+      "loss": 0.0054,
+      "step": 4296
+    },
+    {
+      "epoch": 0.8535107756480286,
+      "grad_norm": 0.46726588405248737,
+      "learning_rate": 5.518531585598452e-07,
+      "loss": 0.0125,
+      "step": 4297
+    },
+    {
+      "epoch": 0.853709405104777,
+      "grad_norm": 0.5320500221062148,
+      "learning_rate": 5.503846893054376e-07,
+      "loss": 0.0147,
+      "step": 4298
+    },
+    {
+      "epoch": 0.8539080345615254,
+      "grad_norm": 0.5026086372784023,
+      "learning_rate": 5.489180626322749e-07,
+      "loss": 0.0147,
+      "step": 4299
+    },
+    {
+      "epoch": 0.854106664018274,
+      "grad_norm": 0.36955112376345584,
+      "learning_rate": 5.474532791476844e-07,
+      "loss": 0.0063,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8543052934750224,
+      "grad_norm": 0.4981679024030091,
+      "learning_rate": 5.459903394582328e-07,
+      "loss": 0.0064,
+      "step": 4301
+    },
+    {
+      "epoch": 0.8545039229317708,
+      "grad_norm": 0.4244213243244062,
+      "learning_rate": 5.445292441697203e-07,
+      "loss": 0.0144,
+      "step": 4302
+    },
+    {
+      "epoch": 0.8547025523885192,
+      "grad_norm": 0.6977278332167148,
+      "learning_rate": 5.430699938871858e-07,
+      "loss": 0.0088,
+      "step": 4303
+    },
+    {
+      "epoch": 0.8549011818452676,
+      "grad_norm": 0.7642388896941835,
+      "learning_rate": 5.416125892149049e-07,
+      "loss": 0.0149,
+      "step": 4304
+    },
+    {
+      "epoch": 0.8550998113020161,
+      "grad_norm": 0.4029155837440189,
+      "learning_rate": 5.401570307563858e-07,
+      "loss": 0.0064,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8552984407587645,
+      "grad_norm": 0.6354016421469559,
+      "learning_rate": 5.387033191143742e-07,
+      "loss": 0.0156,
+      "step": 4306
+    },
+    {
+      "epoch": 0.8554970702155129,
+      "grad_norm": 0.36260476179749174,
+      "learning_rate": 5.372514548908498e-07,
+      "loss": 0.0073,
+      "step": 4307
+    },
+    {
+      "epoch": 0.8556956996722614,
+      "grad_norm": 0.5767576719232735,
+      "learning_rate": 5.358014386870286e-07,
+      "loss": 0.0071,
+      "step": 4308
+    },
+    {
+      "epoch": 0.8558943291290099,
+      "grad_norm": 0.5407542575497629,
+      "learning_rate": 5.343532711033617e-07,
+      "loss": 0.0089,
+      "step": 4309
+    },
+    {
+      "epoch": 0.8560929585857583,
+      "grad_norm": 0.39473103537864135,
+      "learning_rate": 5.329069527395325e-07,
+      "loss": 0.0085,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8562915880425067,
+      "grad_norm": 0.3010851871794071,
+      "learning_rate": 5.314624841944616e-07,
+      "loss": 0.0084,
+      "step": 4311
+    },
+    {
+      "epoch": 0.8564902174992551,
+      "grad_norm": 0.4049809213204187,
+      "learning_rate": 5.300198660663003e-07,
+      "loss": 0.0106,
+      "step": 4312
+    },
+    {
+      "epoch": 0.8566888469560036,
+      "grad_norm": 0.5330916206019888,
+      "learning_rate": 5.28579098952437e-07,
+      "loss": 0.016,
+      "step": 4313
+    },
+    {
+      "epoch": 0.856887476412752,
+      "grad_norm": 0.32358362989327244,
+      "learning_rate": 5.2714018344949e-07,
+      "loss": 0.013,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8570861058695004,
+      "grad_norm": 0.39221129075451266,
+      "learning_rate": 5.257031201533141e-07,
+      "loss": 0.006,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8572847353262489,
+      "grad_norm": 0.4547069807214053,
+      "learning_rate": 5.242679096589959e-07,
+      "loss": 0.0131,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8574833647829974,
+      "grad_norm": 0.24851871977697923,
+      "learning_rate": 5.228345525608536e-07,
+      "loss": 0.007,
+      "step": 4317
+    },
+    {
+      "epoch": 0.8576819942397458,
+      "grad_norm": 0.3943911033364027,
+      "learning_rate": 5.214030494524408e-07,
+      "loss": 0.0136,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8578806236964942,
+      "grad_norm": 0.40090137599123665,
+      "learning_rate": 5.199734009265389e-07,
+      "loss": 0.009,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8580792531532426,
+      "grad_norm": 0.5959732058644688,
+      "learning_rate": 5.18545607575166e-07,
+      "loss": 0.0098,
+      "step": 4320
+    },
+    {
+      "epoch": 0.858277882609991,
+      "grad_norm": 0.6067428479491546,
+      "learning_rate": 5.171196699895687e-07,
+      "loss": 0.0081,
+      "step": 4321
+    },
+    {
+      "epoch": 0.8584765120667395,
+      "grad_norm": 0.46155290089707957,
+      "learning_rate": 5.15695588760226e-07,
+      "loss": 0.0116,
+      "step": 4322
+    },
+    {
+      "epoch": 0.8586751415234879,
+      "grad_norm": 0.4010737738551735,
+      "learning_rate": 5.142733644768511e-07,
+      "loss": 0.0154,
+      "step": 4323
+    },
+    {
+      "epoch": 0.8588737709802363,
+      "grad_norm": 0.2686236162523568,
+      "learning_rate": 5.128529977283824e-07,
+      "loss": 0.0055,
+      "step": 4324
+    },
+    {
+      "epoch": 0.8590724004369849,
+      "grad_norm": 0.618702187000194,
+      "learning_rate": 5.114344891029949e-07,
+      "loss": 0.0114,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8592710298937333,
+      "grad_norm": 0.18789334381194944,
+      "learning_rate": 5.1001783918809e-07,
+      "loss": 0.0039,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8594696593504817,
+      "grad_norm": 0.6465736250291123,
+      "learning_rate": 5.086030485703019e-07,
+      "loss": 0.0084,
+      "step": 4327
+    },
+    {
+      "epoch": 0.8596682888072301,
+      "grad_norm": 0.5719333455155227,
+      "learning_rate": 5.071901178354927e-07,
+      "loss": 0.0094,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8598669182639785,
+      "grad_norm": 0.5000117385352728,
+      "learning_rate": 5.057790475687574e-07,
+      "loss": 0.0065,
+      "step": 4329
+    },
+    {
+      "epoch": 0.860065547720727,
+      "grad_norm": 0.6335598418218465,
+      "learning_rate": 5.043698383544182e-07,
+      "loss": 0.0147,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8602641771774754,
+      "grad_norm": 0.7245941891795274,
+      "learning_rate": 5.029624907760255e-07,
+      "loss": 0.0075,
+      "step": 4331
+    },
+    {
+      "epoch": 0.8604628066342238,
+      "grad_norm": 0.5740508426452544,
+      "learning_rate": 5.015570054163621e-07,
+      "loss": 0.0092,
+      "step": 4332
+    },
+    {
+      "epoch": 0.8606614360909723,
+      "grad_norm": 0.47282095433648685,
+      "learning_rate": 5.001533828574389e-07,
+      "loss": 0.0105,
+      "step": 4333
+    },
+    {
+      "epoch": 0.8608600655477208,
+      "grad_norm": 0.5065183646911192,
+      "learning_rate": 4.987516236804929e-07,
+      "loss": 0.01,
+      "step": 4334
+    },
+    {
+      "epoch": 0.8610586950044692,
+      "grad_norm": 0.6252766191350665,
+      "learning_rate": 4.973517284659923e-07,
+      "loss": 0.0121,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8612573244612176,
+      "grad_norm": 0.8191924217356521,
+      "learning_rate": 4.959536977936313e-07,
+      "loss": 0.0168,
+      "step": 4336
+    },
+    {
+      "epoch": 0.861455953917966,
+      "grad_norm": 0.7702791527849939,
+      "learning_rate": 4.945575322423346e-07,
+      "loss": 0.014,
+      "step": 4337
+    },
+    {
+      "epoch": 0.8616545833747145,
+      "grad_norm": 0.4220973789902732,
+      "learning_rate": 4.931632323902508e-07,
+      "loss": 0.0071,
+      "step": 4338
+    },
+    {
+      "epoch": 0.8618532128314629,
+      "grad_norm": 0.7487513974820313,
+      "learning_rate": 4.917707988147591e-07,
+      "loss": 0.0109,
+      "step": 4339
+    },
+    {
+      "epoch": 0.8620518422882113,
+      "grad_norm": 0.5040477610803438,
+      "learning_rate": 4.903802320924661e-07,
+      "loss": 0.0175,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8622504717449597,
+      "grad_norm": 0.8484536149810432,
+      "learning_rate": 4.889915327992024e-07,
+      "loss": 0.0114,
+      "step": 4341
+    },
+    {
+      "epoch": 0.8624491012017083,
+      "grad_norm": 0.34276135719378503,
+      "learning_rate": 4.876047015100277e-07,
+      "loss": 0.0091,
+      "step": 4342
+    },
+    {
+      "epoch": 0.8626477306584567,
+      "grad_norm": 0.3138227924206721,
+      "learning_rate": 4.862197387992267e-07,
+      "loss": 0.0081,
+      "step": 4343
+    },
+    {
+      "epoch": 0.8628463601152051,
+      "grad_norm": 0.48755996193098844,
+      "learning_rate": 4.848366452403125e-07,
+      "loss": 0.0141,
+      "step": 4344
+    },
+    {
+      "epoch": 0.8630449895719535,
+      "grad_norm": 0.4243160855813772,
+      "learning_rate": 4.834554214060211e-07,
+      "loss": 0.0129,
+      "step": 4345
+    },
+    {
+      "epoch": 0.863243619028702,
+      "grad_norm": 0.28134303568802305,
+      "learning_rate": 4.820760678683168e-07,
+      "loss": 0.0092,
+      "step": 4346
+    },
+    {
+      "epoch": 0.8634422484854504,
+      "grad_norm": 0.9313255602494132,
+      "learning_rate": 4.806985851983892e-07,
+      "loss": 0.0148,
+      "step": 4347
+    },
+    {
+      "epoch": 0.8636408779421988,
+      "grad_norm": 0.33444347069896996,
+      "learning_rate": 4.793229739666505e-07,
+      "loss": 0.0057,
+      "step": 4348
+    },
+    {
+      "epoch": 0.8638395073989472,
+      "grad_norm": 0.4575240882152554,
+      "learning_rate": 4.779492347427422e-07,
+      "loss": 0.0069,
+      "step": 4349
+    },
+    {
+      "epoch": 0.8640381368556957,
+      "grad_norm": 0.25494908087050244,
+      "learning_rate": 4.7657736809552655e-07,
+      "loss": 0.007,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8642367663124442,
+      "grad_norm": 0.674796943258971,
+      "learning_rate": 4.752073745930941e-07,
+      "loss": 0.0126,
+      "step": 4351
+    },
+    {
+      "epoch": 0.8644353957691926,
+      "grad_norm": 0.8848459103510878,
+      "learning_rate": 4.738392548027565e-07,
+      "loss": 0.0094,
+      "step": 4352
+    },
+    {
+      "epoch": 0.864634025225941,
+      "grad_norm": 0.48385172701891577,
+      "learning_rate": 4.724730092910496e-07,
+      "loss": 0.0138,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8648326546826894,
+      "grad_norm": 0.31959578564614055,
+      "learning_rate": 4.7110863862373677e-07,
+      "loss": 0.0059,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8650312841394379,
+      "grad_norm": 0.45707641134122046,
+      "learning_rate": 4.6974614336580014e-07,
+      "loss": 0.0126,
+      "step": 4355
+    },
+    {
+      "epoch": 0.8652299135961863,
+      "grad_norm": 0.24519345456981262,
+      "learning_rate": 4.6838552408145e-07,
+      "loss": 0.006,
+      "step": 4356
+    },
+    {
+      "epoch": 0.8654285430529347,
+      "grad_norm": 0.6553366699161421,
+      "learning_rate": 4.6702678133411505e-07,
+      "loss": 0.0137,
+      "step": 4357
+    },
+    {
+      "epoch": 0.8656271725096832,
+      "grad_norm": 0.6216476033899038,
+      "learning_rate": 4.656699156864508e-07,
+      "loss": 0.0076,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8658258019664317,
+      "grad_norm": 0.9884101474028949,
+      "learning_rate": 4.643149277003345e-07,
+      "loss": 0.0187,
+      "step": 4359
+    },
+    {
+      "epoch": 0.8660244314231801,
+      "grad_norm": 0.3261124525807072,
+      "learning_rate": 4.6296181793686337e-07,
+      "loss": 0.0059,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8662230608799285,
+      "grad_norm": 0.4910131929561135,
+      "learning_rate": 4.616105869563614e-07,
+      "loss": 0.0144,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8664216903366769,
+      "grad_norm": 0.5104540545077465,
+      "learning_rate": 4.602612353183689e-07,
+      "loss": 0.0084,
+      "step": 4362
+    },
+    {
+      "epoch": 0.8666203197934254,
+      "grad_norm": 0.4238162272185062,
+      "learning_rate": 4.589137635816543e-07,
+      "loss": 0.008,
+      "step": 4363
+    },
+    {
+      "epoch": 0.8668189492501738,
+      "grad_norm": 0.5117371152238335,
+      "learning_rate": 4.575681723042014e-07,
+      "loss": 0.0121,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8670175787069222,
+      "grad_norm": 0.4015325998809057,
+      "learning_rate": 4.5622446204321936e-07,
+      "loss": 0.0074,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8672162081636706,
+      "grad_norm": 0.35993039383283,
+      "learning_rate": 4.548826333551382e-07,
+      "loss": 0.0067,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8674148376204192,
+      "grad_norm": 0.503064253155119,
+      "learning_rate": 4.535426867956061e-07,
+      "loss": 0.0073,
+      "step": 4367
+    },
+    {
+      "epoch": 0.8676134670771676,
+      "grad_norm": 0.4651717976831416,
+      "learning_rate": 4.522046229194954e-07,
+      "loss": 0.0075,
+      "step": 4368
+    },
+    {
+      "epoch": 0.867812096533916,
+      "grad_norm": 0.9069952716545454,
+      "learning_rate": 4.508684422808951e-07,
+      "loss": 0.0136,
+      "step": 4369
+    },
+    {
+      "epoch": 0.8680107259906644,
+      "grad_norm": 0.4405186014248274,
+      "learning_rate": 4.4953414543311815e-07,
+      "loss": 0.013,
+      "step": 4370
+    },
+    {
+      "epoch": 0.8682093554474128,
+      "grad_norm": 0.5402377194016759,
+      "learning_rate": 4.4820173292869416e-07,
+      "loss": 0.0143,
+      "step": 4371
+    },
+    {
+      "epoch": 0.8684079849041613,
+      "grad_norm": 0.43013105755484454,
+      "learning_rate": 4.4687120531937357e-07,
+      "loss": 0.0107,
+      "step": 4372
+    },
+    {
+      "epoch": 0.8686066143609097,
+      "grad_norm": 0.4369419022550253,
+      "learning_rate": 4.4554256315612833e-07,
+      "loss": 0.0124,
+      "step": 4373
+    },
+    {
+      "epoch": 0.8688052438176581,
+      "grad_norm": 0.5827970943937599,
+      "learning_rate": 4.4421580698914615e-07,
+      "loss": 0.0091,
+      "step": 4374
+    },
+    {
+      "epoch": 0.8690038732744066,
+      "grad_norm": 0.8284652360245892,
+      "learning_rate": 4.4289093736783695e-07,
+      "loss": 0.0131,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8692025027311551,
+      "grad_norm": 0.5145577977088364,
+      "learning_rate": 4.4156795484082694e-07,
+      "loss": 0.014,
+      "step": 4376
+    },
+    {
+      "epoch": 0.8694011321879035,
+      "grad_norm": 0.24551320324099252,
+      "learning_rate": 4.402468599559606e-07,
+      "loss": 0.0056,
+      "step": 4377
+    },
+    {
+      "epoch": 0.8695997616446519,
+      "grad_norm": 0.44050784980271485,
+      "learning_rate": 4.3892765326030427e-07,
+      "loss": 0.0092,
+      "step": 4378
+    },
+    {
+      "epoch": 0.8697983911014003,
+      "grad_norm": 0.5213423596831668,
+      "learning_rate": 4.376103353001387e-07,
+      "loss": 0.0078,
+      "step": 4379
+    },
+    {
+      "epoch": 0.8699970205581488,
+      "grad_norm": 0.30488209551484446,
+      "learning_rate": 4.3629490662096484e-07,
+      "loss": 0.0071,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8701956500148972,
+      "grad_norm": 0.6465717561681386,
+      "learning_rate": 4.349813677674991e-07,
+      "loss": 0.0066,
+      "step": 4381
+    },
+    {
+      "epoch": 0.8703942794716456,
+      "grad_norm": 0.9699003433798918,
+      "learning_rate": 4.336697192836775e-07,
+      "loss": 0.0155,
+      "step": 4382
+    },
+    {
+      "epoch": 0.870592908928394,
+      "grad_norm": 0.5174827015842891,
+      "learning_rate": 4.323599617126534e-07,
+      "loss": 0.0075,
+      "step": 4383
+    },
+    {
+      "epoch": 0.8707915383851426,
+      "grad_norm": 0.3745007213523484,
+      "learning_rate": 4.3105209559679397e-07,
+      "loss": 0.0083,
+      "step": 4384
+    },
+    {
+      "epoch": 0.870990167841891,
+      "grad_norm": 0.37209009481585975,
+      "learning_rate": 4.2974612147768715e-07,
+      "loss": 0.0075,
+      "step": 4385
+    },
+    {
+      "epoch": 0.8711887972986394,
+      "grad_norm": 0.605148782197854,
+      "learning_rate": 4.284420398961342e-07,
+      "loss": 0.0082,
+      "step": 4386
+    },
+    {
+      "epoch": 0.8713874267553878,
+      "grad_norm": 0.7211710943486761,
+      "learning_rate": 4.2713985139215485e-07,
+      "loss": 0.0083,
+      "step": 4387
+    },
+    {
+      "epoch": 0.8715860562121363,
+      "grad_norm": 0.29150430656637677,
+      "learning_rate": 4.2583955650498276e-07,
+      "loss": 0.0079,
+      "step": 4388
+    },
+    {
+      "epoch": 0.8717846856688847,
+      "grad_norm": 0.6277540412956093,
+      "learning_rate": 4.2454115577307e-07,
+      "loss": 0.0104,
+      "step": 4389
+    },
+    {
+      "epoch": 0.8719833151256331,
+      "grad_norm": 0.8937024962263034,
+      "learning_rate": 4.2324464973408306e-07,
+      "loss": 0.016,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8721819445823815,
+      "grad_norm": 0.22626655374898483,
+      "learning_rate": 4.219500389249026e-07,
+      "loss": 0.0039,
+      "step": 4391
+    },
+    {
+      "epoch": 0.87238057403913,
+      "grad_norm": 0.5483301349143612,
+      "learning_rate": 4.206573238816275e-07,
+      "loss": 0.0116,
+      "step": 4392
+    },
+    {
+      "epoch": 0.8725792034958785,
+      "grad_norm": 0.140657628470087,
+      "learning_rate": 4.193665051395679e-07,
+      "loss": 0.0021,
+      "step": 4393
+    },
+    {
+      "epoch": 0.8727778329526269,
+      "grad_norm": 0.41737224691478453,
+      "learning_rate": 4.180775832332523e-07,
+      "loss": 0.0113,
+      "step": 4394
+    },
+    {
+      "epoch": 0.8729764624093753,
+      "grad_norm": 0.32655825358404184,
+      "learning_rate": 4.1679055869641993e-07,
+      "loss": 0.0078,
+      "step": 4395
+    },
+    {
+      "epoch": 0.8731750918661237,
+      "grad_norm": 0.5392995614964906,
+      "learning_rate": 4.155054320620272e-07,
+      "loss": 0.011,
+      "step": 4396
+    },
+    {
+      "epoch": 0.8733737213228722,
+      "grad_norm": 0.2600770739951475,
+      "learning_rate": 4.1422220386224567e-07,
+      "loss": 0.0055,
+      "step": 4397
+    },
+    {
+      "epoch": 0.8735723507796206,
+      "grad_norm": 0.9550612436854214,
+      "learning_rate": 4.1294087462845576e-07,
+      "loss": 0.0125,
+      "step": 4398
+    },
+    {
+      "epoch": 0.873770980236369,
+      "grad_norm": 0.5668918705649273,
+      "learning_rate": 4.1166144489125703e-07,
+      "loss": 0.0092,
+      "step": 4399
+    },
+    {
+      "epoch": 0.8739696096931175,
+      "grad_norm": 0.3898610540948218,
+      "learning_rate": 4.1038391518045895e-07,
+      "loss": 0.0098,
+      "step": 4400
+    },
+    {
+      "epoch": 0.874168239149866,
+      "grad_norm": 0.542748206533429,
+      "learning_rate": 4.09108286025085e-07,
+      "loss": 0.009,
+      "step": 4401
+    },
+    {
+      "epoch": 0.8743668686066144,
+      "grad_norm": 0.4832034435765608,
+      "learning_rate": 4.0783455795337267e-07,
+      "loss": 0.0096,
+      "step": 4402
+    },
+    {
+      "epoch": 0.8745654980633628,
+      "grad_norm": 0.5202243446046367,
+      "learning_rate": 4.065627314927706e-07,
+      "loss": 0.0109,
+      "step": 4403
+    },
+    {
+      "epoch": 0.8747641275201112,
+      "grad_norm": 0.5784085936698135,
+      "learning_rate": 4.0529280716994246e-07,
+      "loss": 0.0125,
+      "step": 4404
+    },
+    {
+      "epoch": 0.8749627569768597,
+      "grad_norm": 1.1495995355735231,
+      "learning_rate": 4.0402478551076095e-07,
+      "loss": 0.0127,
+      "step": 4405
+    },
+    {
+      "epoch": 0.8751613864336081,
+      "grad_norm": 0.31144818566157473,
+      "learning_rate": 4.027586670403133e-07,
+      "loss": 0.0075,
+      "step": 4406
+    },
+    {
+      "epoch": 0.8753600158903565,
+      "grad_norm": 0.4268316810889361,
+      "learning_rate": 4.0149445228289787e-07,
+      "loss": 0.0082,
+      "step": 4407
+    },
+    {
+      "epoch": 0.8755586453471049,
+      "grad_norm": 0.6936100882597547,
+      "learning_rate": 4.002321417620242e-07,
+      "loss": 0.0143,
+      "step": 4408
+    },
+    {
+      "epoch": 0.8757572748038535,
+      "grad_norm": 0.3286470991413137,
+      "learning_rate": 3.989717360004153e-07,
+      "loss": 0.006,
+      "step": 4409
+    },
+    {
+      "epoch": 0.8759559042606019,
+      "grad_norm": 0.8501238578826438,
+      "learning_rate": 3.9771323552000196e-07,
+      "loss": 0.0141,
+      "step": 4410
+    },
+    {
+      "epoch": 0.8761545337173503,
+      "grad_norm": 0.6051259450991804,
+      "learning_rate": 3.964566408419296e-07,
+      "loss": 0.0109,
+      "step": 4411
+    },
+    {
+      "epoch": 0.8763531631740987,
+      "grad_norm": 0.5210373235659236,
+      "learning_rate": 3.952019524865519e-07,
+      "loss": 0.0086,
+      "step": 4412
+    },
+    {
+      "epoch": 0.8765517926308471,
+      "grad_norm": 1.012680225277243,
+      "learning_rate": 3.939491709734344e-07,
+      "loss": 0.0125,
+      "step": 4413
+    },
+    {
+      "epoch": 0.8767504220875956,
+      "grad_norm": 0.32859384106789974,
+      "learning_rate": 3.926982968213522e-07,
+      "loss": 0.0084,
+      "step": 4414
+    },
+    {
+      "epoch": 0.876949051544344,
+      "grad_norm": 0.2505465396417254,
+      "learning_rate": 3.914493305482914e-07,
+      "loss": 0.0059,
+      "step": 4415
+    },
+    {
+      "epoch": 0.8771476810010924,
+      "grad_norm": 0.542229522925549,
+      "learning_rate": 3.902022726714488e-07,
+      "loss": 0.0134,
+      "step": 4416
+    },
+    {
+      "epoch": 0.877346310457841,
+      "grad_norm": 0.4435250855553033,
+      "learning_rate": 3.889571237072276e-07,
+      "loss": 0.0109,
+      "step": 4417
+    },
+    {
+      "epoch": 0.8775449399145894,
+      "grad_norm": 0.7624348302946883,
+      "learning_rate": 3.877138841712447e-07,
+      "loss": 0.0191,
+      "step": 4418
+    },
+    {
+      "epoch": 0.8777435693713378,
+      "grad_norm": 0.39223360906206584,
+      "learning_rate": 3.8647255457832264e-07,
+      "loss": 0.0108,
+      "step": 4419
+    },
+    {
+      "epoch": 0.8779421988280862,
+      "grad_norm": 0.26266628381114066,
+      "learning_rate": 3.8523313544249653e-07,
+      "loss": 0.0055,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8781408282848346,
+      "grad_norm": 0.6195454555078306,
+      "learning_rate": 3.839956272770068e-07,
+      "loss": 0.0088,
+      "step": 4421
+    },
+    {
+      "epoch": 0.8783394577415831,
+      "grad_norm": 0.6243351595233597,
+      "learning_rate": 3.8276003059430523e-07,
+      "loss": 0.0117,
+      "step": 4422
+    },
+    {
+      "epoch": 0.8785380871983315,
+      "grad_norm": 0.5652958236632496,
+      "learning_rate": 3.8152634590605196e-07,
+      "loss": 0.0199,
+      "step": 4423
+    },
+    {
+      "epoch": 0.8787367166550799,
+      "grad_norm": 0.9261366755651307,
+      "learning_rate": 3.80294573723114e-07,
+      "loss": 0.0146,
+      "step": 4424
+    },
+    {
+      "epoch": 0.8789353461118283,
+      "grad_norm": 0.4162303412607668,
+      "learning_rate": 3.7906471455556537e-07,
+      "loss": 0.0077,
+      "step": 4425
+    },
+    {
+      "epoch": 0.8791339755685769,
+      "grad_norm": 0.47965614259858963,
+      "learning_rate": 3.7783676891269216e-07,
+      "loss": 0.0107,
+      "step": 4426
+    },
+    {
+      "epoch": 0.8793326050253253,
+      "grad_norm": 0.758019511885331,
+      "learning_rate": 3.766107373029837e-07,
+      "loss": 0.0135,
+      "step": 4427
+    },
+    {
+      "epoch": 0.8795312344820737,
+      "grad_norm": 1.122056922868509,
+      "learning_rate": 3.753866202341394e-07,
+      "loss": 0.0115,
+      "step": 4428
+    },
+    {
+      "epoch": 0.8797298639388221,
+      "grad_norm": 0.26469452888733724,
+      "learning_rate": 3.7416441821306325e-07,
+      "loss": 0.0068,
+      "step": 4429
+    },
+    {
+      "epoch": 0.8799284933955706,
+      "grad_norm": 0.7299933289853252,
+      "learning_rate": 3.7294413174587043e-07,
+      "loss": 0.015,
+      "step": 4430
+    },
+    {
+      "epoch": 0.880127122852319,
+      "grad_norm": 0.7634736305853262,
+      "learning_rate": 3.717257613378783e-07,
+      "loss": 0.0093,
+      "step": 4431
+    },
+    {
+      "epoch": 0.8803257523090674,
+      "grad_norm": 0.7735715700253204,
+      "learning_rate": 3.7050930749361335e-07,
+      "loss": 0.0156,
+      "step": 4432
+    },
+    {
+      "epoch": 0.8805243817658158,
+      "grad_norm": 0.5172736405040491,
+      "learning_rate": 3.6929477071680876e-07,
+      "loss": 0.0071,
+      "step": 4433
+    },
+    {
+      "epoch": 0.8807230112225644,
+      "grad_norm": 0.744487659123002,
+      "learning_rate": 3.680821515104016e-07,
+      "loss": 0.0143,
+      "step": 4434
+    },
+    {
+      "epoch": 0.8809216406793128,
+      "grad_norm": 0.5354693790419052,
+      "learning_rate": 3.6687145037653804e-07,
+      "loss": 0.0101,
+      "step": 4435
+    },
+    {
+      "epoch": 0.8811202701360612,
+      "grad_norm": 0.3906157093858662,
+      "learning_rate": 3.656626678165659e-07,
+      "loss": 0.0057,
+      "step": 4436
+    },
+    {
+      "epoch": 0.8813188995928096,
+      "grad_norm": 0.37180422992439127,
+      "learning_rate": 3.6445580433104313e-07,
+      "loss": 0.0073,
+      "step": 4437
+    },
+    {
+      "epoch": 0.881517529049558,
+      "grad_norm": 0.3879523154614315,
+      "learning_rate": 3.632508604197288e-07,
+      "loss": 0.0124,
+      "step": 4438
+    },
+    {
+      "epoch": 0.8817161585063065,
+      "grad_norm": 0.5095632886042696,
+      "learning_rate": 3.6204783658158995e-07,
+      "loss": 0.0117,
+      "step": 4439
+    },
+    {
+      "epoch": 0.8819147879630549,
+      "grad_norm": 0.39433819698390393,
+      "learning_rate": 3.60846733314798e-07,
+      "loss": 0.0089,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8821134174198033,
+      "grad_norm": 0.43438384905622585,
+      "learning_rate": 3.596475511167269e-07,
+      "loss": 0.0061,
+      "step": 4441
+    },
+    {
+      "epoch": 0.8823120468765518,
+      "grad_norm": 0.6305498537007402,
+      "learning_rate": 3.584502904839593e-07,
+      "loss": 0.0126,
+      "step": 4442
+    },
+    {
+      "epoch": 0.8825106763333003,
+      "grad_norm": 0.6097637763944662,
+      "learning_rate": 3.57254951912277e-07,
+      "loss": 0.0137,
+      "step": 4443
+    },
+    {
+      "epoch": 0.8827093057900487,
+      "grad_norm": 0.41692408428569017,
+      "learning_rate": 3.560615358966707e-07,
+      "loss": 0.013,
+      "step": 4444
+    },
+    {
+      "epoch": 0.8829079352467971,
+      "grad_norm": 0.4647122199079791,
+      "learning_rate": 3.548700429313312e-07,
+      "loss": 0.0075,
+      "step": 4445
+    },
+    {
+      "epoch": 0.8831065647035455,
+      "grad_norm": 0.748689177498006,
+      "learning_rate": 3.5368047350965496e-07,
+      "loss": 0.0114,
+      "step": 4446
+    },
+    {
+      "epoch": 0.883305194160294,
+      "grad_norm": 0.8174465249948543,
+      "learning_rate": 3.524928281242429e-07,
+      "loss": 0.0073,
+      "step": 4447
+    },
+    {
+      "epoch": 0.8835038236170424,
+      "grad_norm": 0.4656654252231502,
+      "learning_rate": 3.513071072668961e-07,
+      "loss": 0.0111,
+      "step": 4448
+    },
+    {
+      "epoch": 0.8837024530737908,
+      "grad_norm": 0.5205301107761227,
+      "learning_rate": 3.5012331142862065e-07,
+      "loss": 0.0099,
+      "step": 4449
+    },
+    {
+      "epoch": 0.8839010825305392,
+      "grad_norm": 0.42783686791908027,
+      "learning_rate": 3.4894144109962557e-07,
+      "loss": 0.0114,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8840997119872878,
+      "grad_norm": 0.4787814662703929,
+      "learning_rate": 3.477614967693216e-07,
+      "loss": 0.0086,
+      "step": 4451
+    },
+    {
+      "epoch": 0.8842983414440362,
+      "grad_norm": 0.4314781867922066,
+      "learning_rate": 3.4658347892632337e-07,
+      "loss": 0.0103,
+      "step": 4452
+    },
+    {
+      "epoch": 0.8844969709007846,
+      "grad_norm": 0.35834664121669135,
+      "learning_rate": 3.454073880584463e-07,
+      "loss": 0.0095,
+      "step": 4453
+    },
+    {
+      "epoch": 0.884695600357533,
+      "grad_norm": 0.4202353721736912,
+      "learning_rate": 3.4423322465270914e-07,
+      "loss": 0.0112,
+      "step": 4454
+    },
+    {
+      "epoch": 0.8848942298142815,
+      "grad_norm": 1.639304177182807,
+      "learning_rate": 3.4306098919533013e-07,
+      "loss": 0.0088,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8850928592710299,
+      "grad_norm": 0.5271434662250698,
+      "learning_rate": 3.4189068217173216e-07,
+      "loss": 0.0105,
+      "step": 4456
+    },
+    {
+      "epoch": 0.8852914887277783,
+      "grad_norm": 0.4513582350757716,
+      "learning_rate": 3.407223040665386e-07,
+      "loss": 0.0089,
+      "step": 4457
+    },
+    {
+      "epoch": 0.8854901181845267,
+      "grad_norm": 0.43566762732715747,
+      "learning_rate": 3.395558553635725e-07,
+      "loss": 0.0083,
+      "step": 4458
+    },
+    {
+      "epoch": 0.8856887476412753,
+      "grad_norm": 0.4675844643952693,
+      "learning_rate": 3.383913365458602e-07,
+      "loss": 0.0124,
+      "step": 4459
+    },
+    {
+      "epoch": 0.8858873770980237,
+      "grad_norm": 0.5086115524011803,
+      "learning_rate": 3.3722874809562655e-07,
+      "loss": 0.0085,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8860860065547721,
+      "grad_norm": 0.4995927325240558,
+      "learning_rate": 3.3606809049429976e-07,
+      "loss": 0.0147,
+      "step": 4461
+    },
+    {
+      "epoch": 0.8862846360115205,
+      "grad_norm": 0.563092019434094,
+      "learning_rate": 3.3490936422250486e-07,
+      "loss": 0.0091,
+      "step": 4462
+    },
+    {
+      "epoch": 0.8864832654682689,
+      "grad_norm": 0.3664198660805334,
+      "learning_rate": 3.337525697600713e-07,
+      "loss": 0.0085,
+      "step": 4463
+    },
+    {
+      "epoch": 0.8866818949250174,
+      "grad_norm": 0.5626785718095834,
+      "learning_rate": 3.3259770758602593e-07,
+      "loss": 0.0078,
+      "step": 4464
+    },
+    {
+      "epoch": 0.8868805243817658,
+      "grad_norm": 0.6720037981827449,
+      "learning_rate": 3.314447781785951e-07,
+      "loss": 0.0166,
+      "step": 4465
+    },
+    {
+      "epoch": 0.8870791538385142,
+      "grad_norm": 0.6089564989189392,
+      "learning_rate": 3.302937820152069e-07,
+      "loss": 0.0109,
+      "step": 4466
+    },
+    {
+      "epoch": 0.8872777832952626,
+      "grad_norm": 0.8451481981228195,
+      "learning_rate": 3.291447195724867e-07,
+      "loss": 0.0129,
+      "step": 4467
+    },
+    {
+      "epoch": 0.8874764127520112,
+      "grad_norm": 0.44291299876872686,
+      "learning_rate": 3.2799759132626176e-07,
+      "loss": 0.0075,
+      "step": 4468
+    },
+    {
+      "epoch": 0.8876750422087596,
+      "grad_norm": 0.6004352791066179,
+      "learning_rate": 3.268523977515542e-07,
+      "loss": 0.0122,
+      "step": 4469
+    },
+    {
+      "epoch": 0.887873671665508,
+      "grad_norm": 0.4086550054310082,
+      "learning_rate": 3.257091393225892e-07,
+      "loss": 0.0087,
+      "step": 4470
+    },
+    {
+      "epoch": 0.8880723011222564,
+      "grad_norm": 0.46507471921866933,
+      "learning_rate": 3.245678165127891e-07,
+      "loss": 0.0073,
+      "step": 4471
+    },
+    {
+      "epoch": 0.8882709305790049,
+      "grad_norm": 0.39929592869202446,
+      "learning_rate": 3.234284297947748e-07,
+      "loss": 0.0076,
+      "step": 4472
+    },
+    {
+      "epoch": 0.8884695600357533,
+      "grad_norm": 0.44202911846396886,
+      "learning_rate": 3.222909796403639e-07,
+      "loss": 0.0103,
+      "step": 4473
+    },
+    {
+      "epoch": 0.8886681894925017,
+      "grad_norm": 0.4416813065644919,
+      "learning_rate": 3.211554665205735e-07,
+      "loss": 0.0085,
+      "step": 4474
+    },
+    {
+      "epoch": 0.8888668189492501,
+      "grad_norm": 0.451292002362962,
+      "learning_rate": 3.200218909056185e-07,
+      "loss": 0.0072,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8890654484059987,
+      "grad_norm": 0.39357809600693094,
+      "learning_rate": 3.18890253264913e-07,
+      "loss": 0.0076,
+      "step": 4476
+    },
+    {
+      "epoch": 0.8892640778627471,
+      "grad_norm": 0.599306137463403,
+      "learning_rate": 3.1776055406706474e-07,
+      "loss": 0.0094,
+      "step": 4477
+    },
+    {
+      "epoch": 0.8894627073194955,
+      "grad_norm": 0.3551885948768195,
+      "learning_rate": 3.16632793779883e-07,
+      "loss": 0.0111,
+      "step": 4478
+    },
+    {
+      "epoch": 0.8896613367762439,
+      "grad_norm": 0.303987373083012,
+      "learning_rate": 3.155069728703708e-07,
+      "loss": 0.0064,
+      "step": 4479
+    },
+    {
+      "epoch": 0.8898599662329923,
+      "grad_norm": 0.7886640360675943,
+      "learning_rate": 3.1438309180473083e-07,
+      "loss": 0.0145,
+      "step": 4480
+    },
+    {
+      "epoch": 0.8900585956897408,
+      "grad_norm": 0.3640471787782126,
+      "learning_rate": 3.132611510483591e-07,
+      "loss": 0.0081,
+      "step": 4481
+    },
+    {
+      "epoch": 0.8902572251464892,
+      "grad_norm": 0.4516366521716688,
+      "learning_rate": 3.121411510658512e-07,
+      "loss": 0.0058,
+      "step": 4482
+    },
+    {
+      "epoch": 0.8904558546032376,
+      "grad_norm": 0.5960707874212607,
+      "learning_rate": 3.1102309232099895e-07,
+      "loss": 0.0079,
+      "step": 4483
+    },
+    {
+      "epoch": 0.8906544840599862,
+      "grad_norm": 0.6565899058793593,
+      "learning_rate": 3.09906975276788e-07,
+      "loss": 0.0106,
+      "step": 4484
+    },
+    {
+      "epoch": 0.8908531135167346,
+      "grad_norm": 0.43991379023445737,
+      "learning_rate": 3.087928003954027e-07,
+      "loss": 0.0079,
+      "step": 4485
+    },
+    {
+      "epoch": 0.891051742973483,
+      "grad_norm": 0.5593853069391717,
+      "learning_rate": 3.0768056813821943e-07,
+      "loss": 0.0135,
+      "step": 4486
+    },
+    {
+      "epoch": 0.8912503724302314,
+      "grad_norm": 0.3950431747898158,
+      "learning_rate": 3.0657027896581537e-07,
+      "loss": 0.0094,
+      "step": 4487
+    },
+    {
+      "epoch": 0.8914490018869798,
+      "grad_norm": 0.25409553815336716,
+      "learning_rate": 3.054619333379577e-07,
+      "loss": 0.0034,
+      "step": 4488
+    },
+    {
+      "epoch": 0.8916476313437283,
+      "grad_norm": 0.5330841755954928,
+      "learning_rate": 3.0435553171361207e-07,
+      "loss": 0.0072,
+      "step": 4489
+    },
+    {
+      "epoch": 0.8918462608004767,
+      "grad_norm": 0.6056435536748148,
+      "learning_rate": 3.032510745509393e-07,
+      "loss": 0.0125,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8920448902572251,
+      "grad_norm": 0.40977754735724337,
+      "learning_rate": 3.021485623072923e-07,
+      "loss": 0.0091,
+      "step": 4491
+    },
+    {
+      "epoch": 0.8922435197139735,
+      "grad_norm": 0.5596491771836665,
+      "learning_rate": 3.0104799543922146e-07,
+      "loss": 0.0105,
+      "step": 4492
+    },
+    {
+      "epoch": 0.8924421491707221,
+      "grad_norm": 0.4786492106128223,
+      "learning_rate": 2.999493744024701e-07,
+      "loss": 0.0083,
+      "step": 4493
+    },
+    {
+      "epoch": 0.8926407786274705,
+      "grad_norm": 0.9836960055347553,
+      "learning_rate": 2.988526996519764e-07,
+      "loss": 0.018,
+      "step": 4494
+    },
+    {
+      "epoch": 0.8928394080842189,
+      "grad_norm": 0.6957875722193689,
+      "learning_rate": 2.97757971641871e-07,
+      "loss": 0.0072,
+      "step": 4495
+    },
+    {
+      "epoch": 0.8930380375409673,
+      "grad_norm": 0.7150594587760443,
+      "learning_rate": 2.966651908254814e-07,
+      "loss": 0.0151,
+      "step": 4496
+    },
+    {
+      "epoch": 0.8932366669977158,
+      "grad_norm": 0.9452942090466192,
+      "learning_rate": 2.955743576553266e-07,
+      "loss": 0.011,
+      "step": 4497
+    },
+    {
+      "epoch": 0.8934352964544642,
+      "grad_norm": 0.4147804118671288,
+      "learning_rate": 2.944854725831181e-07,
+      "loss": 0.0086,
+      "step": 4498
+    },
+    {
+      "epoch": 0.8936339259112126,
+      "grad_norm": 0.6444712673599818,
+      "learning_rate": 2.93398536059763e-07,
+      "loss": 0.0116,
+      "step": 4499
+    },
+    {
+      "epoch": 0.893832555367961,
+      "grad_norm": 0.4545175271742511,
+      "learning_rate": 2.923135485353618e-07,
+      "loss": 0.0091,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8940311848247096,
+      "grad_norm": 0.435024614613689,
+      "learning_rate": 2.9123051045920503e-07,
+      "loss": 0.0099,
+      "step": 4501
+    },
+    {
+      "epoch": 0.894229814281458,
+      "grad_norm": 0.5137299220895429,
+      "learning_rate": 2.9014942227977907e-07,
+      "loss": 0.0114,
+      "step": 4502
+    },
+    {
+      "epoch": 0.8944284437382064,
+      "grad_norm": 0.5936552009342531,
+      "learning_rate": 2.8907028444476017e-07,
+      "loss": 0.0147,
+      "step": 4503
+    },
+    {
+      "epoch": 0.8946270731949548,
+      "grad_norm": 0.6272540958685057,
+      "learning_rate": 2.879930974010198e-07,
+      "loss": 0.0118,
+      "step": 4504
+    },
+    {
+      "epoch": 0.8948257026517032,
+      "grad_norm": 0.626455749063314,
+      "learning_rate": 2.8691786159461776e-07,
+      "loss": 0.0086,
+      "step": 4505
+    },
+    {
+      "epoch": 0.8950243321084517,
+      "grad_norm": 0.4727682477458918,
+      "learning_rate": 2.858445774708096e-07,
+      "loss": 0.0125,
+      "step": 4506
+    },
+    {
+      "epoch": 0.8952229615652001,
+      "grad_norm": 0.42330205421234,
+      "learning_rate": 2.8477324547404204e-07,
+      "loss": 0.0131,
+      "step": 4507
+    },
+    {
+      "epoch": 0.8954215910219485,
+      "grad_norm": 0.8016337669641905,
+      "learning_rate": 2.837038660479508e-07,
+      "loss": 0.0087,
+      "step": 4508
+    },
+    {
+      "epoch": 0.8956202204786969,
+      "grad_norm": 0.4415565035557052,
+      "learning_rate": 2.826364396353659e-07,
+      "loss": 0.0127,
+      "step": 4509
+    },
+    {
+      "epoch": 0.8958188499354455,
+      "grad_norm": 0.5118589639856184,
+      "learning_rate": 2.815709666783073e-07,
+      "loss": 0.0076,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8960174793921939,
+      "grad_norm": 0.7602872016834586,
+      "learning_rate": 2.805074476179864e-07,
+      "loss": 0.0098,
+      "step": 4511
+    },
+    {
+      "epoch": 0.8962161088489423,
+      "grad_norm": 0.29573595687294996,
+      "learning_rate": 2.7944588289480436e-07,
+      "loss": 0.0071,
+      "step": 4512
+    },
+    {
+      "epoch": 0.8964147383056907,
+      "grad_norm": 0.6148007368180166,
+      "learning_rate": 2.7838627294835553e-07,
+      "loss": 0.0177,
+      "step": 4513
+    },
+    {
+      "epoch": 0.8966133677624392,
+      "grad_norm": 0.3086576744890267,
+      "learning_rate": 2.7732861821742285e-07,
+      "loss": 0.0037,
+      "step": 4514
+    },
+    {
+      "epoch": 0.8968119972191876,
+      "grad_norm": 0.33506300714770954,
+      "learning_rate": 2.762729191399799e-07,
+      "loss": 0.0078,
+      "step": 4515
+    },
+    {
+      "epoch": 0.897010626675936,
+      "grad_norm": 0.7280224457903677,
+      "learning_rate": 2.752191761531908e-07,
+      "loss": 0.0132,
+      "step": 4516
+    },
+    {
+      "epoch": 0.8972092561326844,
+      "grad_norm": 0.902157920426683,
+      "learning_rate": 2.7416738969340884e-07,
+      "loss": 0.0157,
+      "step": 4517
+    },
+    {
+      "epoch": 0.897407885589433,
+      "grad_norm": 1.5039398413544367,
+      "learning_rate": 2.7311756019617886e-07,
+      "loss": 0.0096,
+      "step": 4518
+    },
+    {
+      "epoch": 0.8976065150461814,
+      "grad_norm": 0.43159755127163674,
+      "learning_rate": 2.720696880962331e-07,
+      "loss": 0.0142,
+      "step": 4519
+    },
+    {
+      "epoch": 0.8978051445029298,
+      "grad_norm": 0.43080340115106225,
+      "learning_rate": 2.710237738274951e-07,
+      "loss": 0.0147,
+      "step": 4520
+    },
+    {
+      "epoch": 0.8980037739596782,
+      "grad_norm": 0.7326367128578075,
+      "learning_rate": 2.699798178230772e-07,
+      "loss": 0.0189,
+      "step": 4521
+    },
+    {
+      "epoch": 0.8982024034164267,
+      "grad_norm": 0.43562861431632444,
+      "learning_rate": 2.6893782051527873e-07,
+      "loss": 0.011,
+      "step": 4522
+    },
+    {
+      "epoch": 0.8984010328731751,
+      "grad_norm": 0.3484366560219996,
+      "learning_rate": 2.6789778233559214e-07,
+      "loss": 0.0075,
+      "step": 4523
+    },
+    {
+      "epoch": 0.8985996623299235,
+      "grad_norm": 0.45427660263747915,
+      "learning_rate": 2.6685970371469414e-07,
+      "loss": 0.0057,
+      "step": 4524
+    },
+    {
+      "epoch": 0.8987982917866719,
+      "grad_norm": 0.34514097610134814,
+      "learning_rate": 2.658235850824531e-07,
+      "loss": 0.009,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8989969212434203,
+      "grad_norm": 0.9536404644314742,
+      "learning_rate": 2.647894268679252e-07,
+      "loss": 0.0058,
+      "step": 4526
+    },
+    {
+      "epoch": 0.8991955507001689,
+      "grad_norm": 0.679047679167587,
+      "learning_rate": 2.637572294993529e-07,
+      "loss": 0.0158,
+      "step": 4527
+    },
+    {
+      "epoch": 0.8993941801569173,
+      "grad_norm": 0.48771424479973163,
+      "learning_rate": 2.627269934041693e-07,
+      "loss": 0.0082,
+      "step": 4528
+    },
+    {
+      "epoch": 0.8995928096136657,
+      "grad_norm": 0.8876585963703878,
+      "learning_rate": 2.6169871900899367e-07,
+      "loss": 0.0129,
+      "step": 4529
+    },
+    {
+      "epoch": 0.8997914390704141,
+      "grad_norm": 0.42451999038659616,
+      "learning_rate": 2.606724067396338e-07,
+      "loss": 0.0106,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8999900685271626,
+      "grad_norm": 0.5490668531272829,
+      "learning_rate": 2.59648057021083e-07,
+      "loss": 0.0152,
+      "step": 4531
+    },
+    {
+      "epoch": 0.900188697983911,
+      "grad_norm": 0.5470093186008119,
+      "learning_rate": 2.5862567027752526e-07,
+      "loss": 0.0094,
+      "step": 4532
+    },
+    {
+      "epoch": 0.9003873274406594,
+      "grad_norm": 0.6398893462509314,
+      "learning_rate": 2.576052469323298e-07,
+      "loss": 0.0116,
+      "step": 4533
+    },
+    {
+      "epoch": 0.9005859568974078,
+      "grad_norm": 0.4022552451748213,
+      "learning_rate": 2.5658678740805186e-07,
+      "loss": 0.0077,
+      "step": 4534
+    },
+    {
+      "epoch": 0.9007845863541564,
+      "grad_norm": 0.39557786288450486,
+      "learning_rate": 2.555702921264358e-07,
+      "loss": 0.0088,
+      "step": 4535
+    },
+    {
+      "epoch": 0.9009832158109048,
+      "grad_norm": 1.0099485838483995,
+      "learning_rate": 2.545557615084099e-07,
+      "loss": 0.0117,
+      "step": 4536
+    },
+    {
+      "epoch": 0.9011818452676532,
+      "grad_norm": 0.6052324571559026,
+      "learning_rate": 2.5354319597409194e-07,
+      "loss": 0.0095,
+      "step": 4537
+    },
+    {
+      "epoch": 0.9013804747244016,
+      "grad_norm": 0.4649008984948733,
+      "learning_rate": 2.525325959427821e-07,
+      "loss": 0.0056,
+      "step": 4538
+    },
+    {
+      "epoch": 0.9015791041811501,
+      "grad_norm": 0.4320126986367103,
+      "learning_rate": 2.515239618329701e-07,
+      "loss": 0.0102,
+      "step": 4539
+    },
+    {
+      "epoch": 0.9017777336378985,
+      "grad_norm": 0.3576353257805511,
+      "learning_rate": 2.505172940623313e-07,
+      "loss": 0.0062,
+      "step": 4540
+    },
+    {
+      "epoch": 0.9019763630946469,
+      "grad_norm": 0.3318197981973977,
+      "learning_rate": 2.495125930477238e-07,
+      "loss": 0.01,
+      "step": 4541
+    },
+    {
+      "epoch": 0.9021749925513953,
+      "grad_norm": 0.4512411192015429,
+      "learning_rate": 2.485098592051949e-07,
+      "loss": 0.0048,
+      "step": 4542
+    },
+    {
+      "epoch": 0.9023736220081439,
+      "grad_norm": 0.3890900313378956,
+      "learning_rate": 2.475090929499746e-07,
+      "loss": 0.0101,
+      "step": 4543
+    },
+    {
+      "epoch": 0.9025722514648923,
+      "grad_norm": 0.43260373969735505,
+      "learning_rate": 2.465102946964798e-07,
+      "loss": 0.0148,
+      "step": 4544
+    },
+    {
+      "epoch": 0.9027708809216407,
+      "grad_norm": 0.586975403052094,
+      "learning_rate": 2.455134648583124e-07,
+      "loss": 0.0123,
+      "step": 4545
+    },
+    {
+      "epoch": 0.9029695103783891,
+      "grad_norm": 0.4911950176274392,
+      "learning_rate": 2.4451860384825663e-07,
+      "loss": 0.011,
+      "step": 4546
+    },
+    {
+      "epoch": 0.9031681398351376,
+      "grad_norm": 0.389644189362611,
+      "learning_rate": 2.4352571207828577e-07,
+      "loss": 0.011,
+      "step": 4547
+    },
+    {
+      "epoch": 0.903366769291886,
+      "grad_norm": 0.43830143444725367,
+      "learning_rate": 2.425347899595537e-07,
+      "loss": 0.0056,
+      "step": 4548
+    },
+    {
+      "epoch": 0.9035653987486344,
+      "grad_norm": 0.6957717254966186,
+      "learning_rate": 2.415458379024005e-07,
+      "loss": 0.0163,
+      "step": 4549
+    },
+    {
+      "epoch": 0.9037640282053828,
+      "grad_norm": 0.4740762518443394,
+      "learning_rate": 2.405588563163519e-07,
+      "loss": 0.0146,
+      "step": 4550
+    },
+    {
+      "epoch": 0.9039626576621312,
+      "grad_norm": 0.6021255205184528,
+      "learning_rate": 2.395738456101132e-07,
+      "loss": 0.0124,
+      "step": 4551
+    },
+    {
+      "epoch": 0.9041612871188798,
+      "grad_norm": 0.741752144803565,
+      "learning_rate": 2.3859080619157925e-07,
+      "loss": 0.0125,
+      "step": 4552
+    },
+    {
+      "epoch": 0.9043599165756282,
+      "grad_norm": 0.42927765078239116,
+      "learning_rate": 2.376097384678233e-07,
+      "loss": 0.0156,
+      "step": 4553
+    },
+    {
+      "epoch": 0.9045585460323766,
+      "grad_norm": 0.6840710533669623,
+      "learning_rate": 2.3663064284510594e-07,
+      "loss": 0.0157,
+      "step": 4554
+    },
+    {
+      "epoch": 0.904757175489125,
+      "grad_norm": 0.62982881490239,
+      "learning_rate": 2.356535197288684e-07,
+      "loss": 0.0129,
+      "step": 4555
+    },
+    {
+      "epoch": 0.9049558049458735,
+      "grad_norm": 0.35044902975137593,
+      "learning_rate": 2.3467836952373756e-07,
+      "loss": 0.0073,
+      "step": 4556
+    },
+    {
+      "epoch": 0.9051544344026219,
+      "grad_norm": 0.5392210976679882,
+      "learning_rate": 2.3370519263352264e-07,
+      "loss": 0.0094,
+      "step": 4557
+    },
+    {
+      "epoch": 0.9053530638593703,
+      "grad_norm": 0.48828299525425184,
+      "learning_rate": 2.327339894612135e-07,
+      "loss": 0.0153,
+      "step": 4558
+    },
+    {
+      "epoch": 0.9055516933161187,
+      "grad_norm": 0.4056484456009865,
+      "learning_rate": 2.3176476040898566e-07,
+      "loss": 0.0135,
+      "step": 4559
+    },
+    {
+      "epoch": 0.9057503227728673,
+      "grad_norm": 0.6301698090976487,
+      "learning_rate": 2.3079750587819527e-07,
+      "loss": 0.0113,
+      "step": 4560
+    },
+    {
+      "epoch": 0.9059489522296157,
+      "grad_norm": 0.48414953276680756,
+      "learning_rate": 2.2983222626938196e-07,
+      "loss": 0.0165,
+      "step": 4561
+    },
+    {
+      "epoch": 0.9061475816863641,
+      "grad_norm": 0.44003626001855367,
+      "learning_rate": 2.288689219822665e-07,
+      "loss": 0.0124,
+      "step": 4562
+    },
+    {
+      "epoch": 0.9063462111431125,
+      "grad_norm": 0.6813734049579375,
+      "learning_rate": 2.2790759341575208e-07,
+      "loss": 0.0116,
+      "step": 4563
+    },
+    {
+      "epoch": 0.906544840599861,
+      "grad_norm": 0.5755670216305095,
+      "learning_rate": 2.2694824096792522e-07,
+      "loss": 0.0124,
+      "step": 4564
+    },
+    {
+      "epoch": 0.9067434700566094,
+      "grad_norm": 0.40263299299179744,
+      "learning_rate": 2.259908650360515e-07,
+      "loss": 0.0122,
+      "step": 4565
+    },
+    {
+      "epoch": 0.9069420995133578,
+      "grad_norm": 0.6401801441669921,
+      "learning_rate": 2.2503546601657988e-07,
+      "loss": 0.0105,
+      "step": 4566
+    },
+    {
+      "epoch": 0.9071407289701062,
+      "grad_norm": 0.785747890099103,
+      "learning_rate": 2.2408204430514003e-07,
+      "loss": 0.0169,
+      "step": 4567
+    },
+    {
+      "epoch": 0.9073393584268546,
+      "grad_norm": 0.4264636528520051,
+      "learning_rate": 2.2313060029654276e-07,
+      "loss": 0.0059,
+      "step": 4568
+    },
+    {
+      "epoch": 0.9075379878836032,
+      "grad_norm": 0.42739671235798343,
+      "learning_rate": 2.2218113438478074e-07,
+      "loss": 0.0083,
+      "step": 4569
+    },
+    {
+      "epoch": 0.9077366173403516,
+      "grad_norm": 0.40779650096228554,
+      "learning_rate": 2.2123364696302553e-07,
+      "loss": 0.0077,
+      "step": 4570
+    },
+    {
+      "epoch": 0.9079352467971,
+      "grad_norm": 0.39853271354961034,
+      "learning_rate": 2.2028813842363272e-07,
+      "loss": 0.0107,
+      "step": 4571
+    },
+    {
+      "epoch": 0.9081338762538484,
+      "grad_norm": 0.33762970470811593,
+      "learning_rate": 2.1934460915813416e-07,
+      "loss": 0.0075,
+      "step": 4572
+    },
+    {
+      "epoch": 0.9083325057105969,
+      "grad_norm": 0.5183895697114581,
+      "learning_rate": 2.1840305955724561e-07,
+      "loss": 0.0096,
+      "step": 4573
+    },
+    {
+      "epoch": 0.9085311351673453,
+      "grad_norm": 0.48886933649478814,
+      "learning_rate": 2.1746349001086187e-07,
+      "loss": 0.0061,
+      "step": 4574
+    },
+    {
+      "epoch": 0.9087297646240937,
+      "grad_norm": 0.638424883774554,
+      "learning_rate": 2.1652590090805725e-07,
+      "loss": 0.0085,
+      "step": 4575
+    },
+    {
+      "epoch": 0.9089283940808421,
+      "grad_norm": 0.47399133102593094,
+      "learning_rate": 2.155902926370873e-07,
+      "loss": 0.0102,
+      "step": 4576
+    },
+    {
+      "epoch": 0.9091270235375907,
+      "grad_norm": 0.8460449737384751,
+      "learning_rate": 2.1465666558538544e-07,
+      "loss": 0.0103,
+      "step": 4577
+    },
+    {
+      "epoch": 0.9093256529943391,
+      "grad_norm": 0.5377971540122163,
+      "learning_rate": 2.1372502013956687e-07,
+      "loss": 0.0114,
+      "step": 4578
+    },
+    {
+      "epoch": 0.9095242824510875,
+      "grad_norm": 0.4809164280538038,
+      "learning_rate": 2.127953566854235e-07,
+      "loss": 0.0105,
+      "step": 4579
+    },
+    {
+      "epoch": 0.9097229119078359,
+      "grad_norm": 0.2805309828097,
+      "learning_rate": 2.1186767560792964e-07,
+      "loss": 0.0079,
+      "step": 4580
+    },
+    {
+      "epoch": 0.9099215413645844,
+      "grad_norm": 0.43230047439300484,
+      "learning_rate": 2.1094197729123577e-07,
+      "loss": 0.012,
+      "step": 4581
+    },
+    {
+      "epoch": 0.9101201708213328,
+      "grad_norm": 0.5605835051911726,
+      "learning_rate": 2.100182621186736e-07,
+      "loss": 0.0159,
+      "step": 4582
+    },
+    {
+      "epoch": 0.9103188002780812,
+      "grad_norm": 0.5202163319138345,
+      "learning_rate": 2.090965304727527e-07,
+      "loss": 0.0096,
+      "step": 4583
+    },
+    {
+      "epoch": 0.9105174297348296,
+      "grad_norm": 0.4097715169980986,
+      "learning_rate": 2.0817678273515996e-07,
+      "loss": 0.0107,
+      "step": 4584
+    },
+    {
+      "epoch": 0.9107160591915782,
+      "grad_norm": 0.8758598360143836,
+      "learning_rate": 2.072590192867635e-07,
+      "loss": 0.0103,
+      "step": 4585
+    },
+    {
+      "epoch": 0.9109146886483266,
+      "grad_norm": 0.5918227290134394,
+      "learning_rate": 2.0634324050760658e-07,
+      "loss": 0.0122,
+      "step": 4586
+    },
+    {
+      "epoch": 0.911113318105075,
+      "grad_norm": 0.49732322398821266,
+      "learning_rate": 2.0542944677691467e-07,
+      "loss": 0.0112,
+      "step": 4587
+    },
+    {
+      "epoch": 0.9113119475618234,
+      "grad_norm": 0.7732911114954113,
+      "learning_rate": 2.0451763847308626e-07,
+      "loss": 0.0122,
+      "step": 4588
+    },
+    {
+      "epoch": 0.9115105770185719,
+      "grad_norm": 0.6007278204049679,
+      "learning_rate": 2.036078159737015e-07,
+      "loss": 0.0139,
+      "step": 4589
+    },
+    {
+      "epoch": 0.9117092064753203,
+      "grad_norm": 1.284947054234802,
+      "learning_rate": 2.026999796555179e-07,
+      "loss": 0.0237,
+      "step": 4590
+    },
+    {
+      "epoch": 0.9119078359320687,
+      "grad_norm": 0.567152964186292,
+      "learning_rate": 2.0179412989446756e-07,
+      "loss": 0.0088,
+      "step": 4591
+    },
+    {
+      "epoch": 0.9121064653888171,
+      "grad_norm": 0.38390704525282005,
+      "learning_rate": 2.0089026706566372e-07,
+      "loss": 0.0067,
+      "step": 4592
+    },
+    {
+      "epoch": 0.9123050948455655,
+      "grad_norm": 0.6975714046203888,
+      "learning_rate": 1.999883915433948e-07,
+      "loss": 0.017,
+      "step": 4593
+    },
+    {
+      "epoch": 0.9125037243023141,
+      "grad_norm": 0.5217159598877467,
+      "learning_rate": 1.9908850370112476e-07,
+      "loss": 0.0149,
+      "step": 4594
+    },
+    {
+      "epoch": 0.9127023537590625,
+      "grad_norm": 0.5062818955759204,
+      "learning_rate": 1.9819060391149837e-07,
+      "loss": 0.0045,
+      "step": 4595
+    },
+    {
+      "epoch": 0.9129009832158109,
+      "grad_norm": 0.30318092705281424,
+      "learning_rate": 1.9729469254633425e-07,
+      "loss": 0.0063,
+      "step": 4596
+    },
+    {
+      "epoch": 0.9130996126725593,
+      "grad_norm": 0.2722804564584692,
+      "learning_rate": 1.9640076997662848e-07,
+      "loss": 0.0067,
+      "step": 4597
+    },
+    {
+      "epoch": 0.9132982421293078,
+      "grad_norm": 0.4823650756559266,
+      "learning_rate": 1.955088365725527e-07,
+      "loss": 0.0092,
+      "step": 4598
+    },
+    {
+      "epoch": 0.9134968715860562,
+      "grad_norm": 0.45532989870452495,
+      "learning_rate": 1.9461889270345645e-07,
+      "loss": 0.011,
+      "step": 4599
+    },
+    {
+      "epoch": 0.9136955010428046,
+      "grad_norm": 0.6373971983656296,
+      "learning_rate": 1.9373093873786497e-07,
+      "loss": 0.0084,
+      "step": 4600
+    },
+    {
+      "epoch": 0.913894130499553,
+      "grad_norm": 0.491770676105079,
+      "learning_rate": 1.9284497504347854e-07,
+      "loss": 0.006,
+      "step": 4601
+    },
+    {
+      "epoch": 0.9140927599563016,
+      "grad_norm": 0.6167707314340485,
+      "learning_rate": 1.9196100198717427e-07,
+      "loss": 0.0161,
+      "step": 4602
+    },
+    {
+      "epoch": 0.91429138941305,
+      "grad_norm": 0.462717621588757,
+      "learning_rate": 1.9107901993500322e-07,
+      "loss": 0.0157,
+      "step": 4603
+    },
+    {
+      "epoch": 0.9144900188697984,
+      "grad_norm": 0.17662504554317196,
+      "learning_rate": 1.9019902925219548e-07,
+      "loss": 0.004,
+      "step": 4604
+    },
+    {
+      "epoch": 0.9146886483265468,
+      "grad_norm": 0.49148437356429236,
+      "learning_rate": 1.893210303031523e-07,
+      "loss": 0.0121,
+      "step": 4605
+    },
+    {
+      "epoch": 0.9148872777832953,
+      "grad_norm": 0.7776087485666608,
+      "learning_rate": 1.8844502345145233e-07,
+      "loss": 0.0096,
+      "step": 4606
+    },
+    {
+      "epoch": 0.9150859072400437,
+      "grad_norm": 0.5332572925885097,
+      "learning_rate": 1.8757100905985094e-07,
+      "loss": 0.0055,
+      "step": 4607
+    },
+    {
+      "epoch": 0.9152845366967921,
+      "grad_norm": 0.32966084466808204,
+      "learning_rate": 1.8669898749027472e-07,
+      "loss": 0.0094,
+      "step": 4608
+    },
+    {
+      "epoch": 0.9154831661535405,
+      "grad_norm": 0.3990945821247028,
+      "learning_rate": 1.8582895910382813e-07,
+      "loss": 0.0125,
+      "step": 4609
+    },
+    {
+      "epoch": 0.915681795610289,
+      "grad_norm": 0.44485780478520864,
+      "learning_rate": 1.8496092426078805e-07,
+      "loss": 0.0074,
+      "step": 4610
+    },
+    {
+      "epoch": 0.9158804250670375,
+      "grad_norm": 0.558341054302273,
+      "learning_rate": 1.8409488332060799e-07,
+      "loss": 0.0115,
+      "step": 4611
+    },
+    {
+      "epoch": 0.9160790545237859,
+      "grad_norm": 0.37536643210218906,
+      "learning_rate": 1.8323083664191333e-07,
+      "loss": 0.0057,
+      "step": 4612
+    },
+    {
+      "epoch": 0.9162776839805343,
+      "grad_norm": 0.7931572433347439,
+      "learning_rate": 1.823687845825056e-07,
+      "loss": 0.0135,
+      "step": 4613
+    },
+    {
+      "epoch": 0.9164763134372828,
+      "grad_norm": 0.5130348406821796,
+      "learning_rate": 1.8150872749935989e-07,
+      "loss": 0.017,
+      "step": 4614
+    },
+    {
+      "epoch": 0.9166749428940312,
+      "grad_norm": 0.5767895711884279,
+      "learning_rate": 1.8065066574862455e-07,
+      "loss": 0.0193,
+      "step": 4615
+    },
+    {
+      "epoch": 0.9168735723507796,
+      "grad_norm": 0.5029366482371473,
+      "learning_rate": 1.7979459968562317e-07,
+      "loss": 0.0144,
+      "step": 4616
+    },
+    {
+      "epoch": 0.917072201807528,
+      "grad_norm": 0.49339222260197624,
+      "learning_rate": 1.7894052966485053e-07,
+      "loss": 0.0069,
+      "step": 4617
+    },
+    {
+      "epoch": 0.9172708312642764,
+      "grad_norm": 0.4827004588018792,
+      "learning_rate": 1.7808845603997594e-07,
+      "loss": 0.0089,
+      "step": 4618
+    },
+    {
+      "epoch": 0.917469460721025,
+      "grad_norm": 0.48201824114443204,
+      "learning_rate": 1.772383791638438e-07,
+      "loss": 0.0117,
+      "step": 4619
+    },
+    {
+      "epoch": 0.9176680901777734,
+      "grad_norm": 0.42119151357623763,
+      "learning_rate": 1.7639029938846808e-07,
+      "loss": 0.0152,
+      "step": 4620
+    },
+    {
+      "epoch": 0.9178667196345218,
+      "grad_norm": 0.42340226485526916,
+      "learning_rate": 1.755442170650401e-07,
+      "loss": 0.0062,
+      "step": 4621
+    },
+    {
+      "epoch": 0.9180653490912702,
+      "grad_norm": 0.42028816533360497,
+      "learning_rate": 1.747001325439196e-07,
+      "loss": 0.0109,
+      "step": 4622
+    },
+    {
+      "epoch": 0.9182639785480187,
+      "grad_norm": 0.37252045119966465,
+      "learning_rate": 1.7385804617464308e-07,
+      "loss": 0.0132,
+      "step": 4623
+    },
+    {
+      "epoch": 0.9184626080047671,
+      "grad_norm": 0.6415858988525973,
+      "learning_rate": 1.730179583059155e-07,
+      "loss": 0.014,
+      "step": 4624
+    },
+    {
+      "epoch": 0.9186612374615155,
+      "grad_norm": 0.6620697719569957,
+      "learning_rate": 1.7217986928561803e-07,
+      "loss": 0.0073,
+      "step": 4625
+    },
+    {
+      "epoch": 0.9188598669182639,
+      "grad_norm": 0.534573318706696,
+      "learning_rate": 1.7134377946080193e-07,
+      "loss": 0.0085,
+      "step": 4626
+    },
+    {
+      "epoch": 0.9190584963750125,
+      "grad_norm": 0.4937748956041207,
+      "learning_rate": 1.7050968917769139e-07,
+      "loss": 0.0108,
+      "step": 4627
+    },
+    {
+      "epoch": 0.9192571258317609,
+      "grad_norm": 0.7534757131719959,
+      "learning_rate": 1.6967759878168233e-07,
+      "loss": 0.0149,
+      "step": 4628
+    },
+    {
+      "epoch": 0.9194557552885093,
+      "grad_norm": 0.48683536600548843,
+      "learning_rate": 1.688475086173419e-07,
+      "loss": 0.0071,
+      "step": 4629
+    },
+    {
+      "epoch": 0.9196543847452577,
+      "grad_norm": 0.5205814438223413,
+      "learning_rate": 1.6801941902841068e-07,
+      "loss": 0.0133,
+      "step": 4630
+    },
+    {
+      "epoch": 0.9198530142020062,
+      "grad_norm": 0.37512216595028314,
+      "learning_rate": 1.6719333035779827e-07,
+      "loss": 0.0075,
+      "step": 4631
+    },
+    {
+      "epoch": 0.9200516436587546,
+      "grad_norm": 0.5986458365684142,
+      "learning_rate": 1.6636924294758828e-07,
+      "loss": 0.0095,
+      "step": 4632
+    },
+    {
+      "epoch": 0.920250273115503,
+      "grad_norm": 0.5772681496058722,
+      "learning_rate": 1.655471571390349e-07,
+      "loss": 0.0083,
+      "step": 4633
+    },
+    {
+      "epoch": 0.9204489025722514,
+      "grad_norm": 0.29855689621547776,
+      "learning_rate": 1.6472707327256198e-07,
+      "loss": 0.0048,
+      "step": 4634
+    },
+    {
+      "epoch": 0.9206475320289998,
+      "grad_norm": 0.5804263227677894,
+      "learning_rate": 1.639089916877662e-07,
+      "loss": 0.0156,
+      "step": 4635
+    },
+    {
+      "epoch": 0.9208461614857484,
+      "grad_norm": 0.47922034362262383,
+      "learning_rate": 1.6309291272341377e-07,
+      "loss": 0.0108,
+      "step": 4636
+    },
+    {
+      "epoch": 0.9210447909424968,
+      "grad_norm": 0.3263256249555732,
+      "learning_rate": 1.622788367174427e-07,
+      "loss": 0.0058,
+      "step": 4637
+    },
+    {
+      "epoch": 0.9212434203992452,
+      "grad_norm": 0.6192492406914758,
+      "learning_rate": 1.6146676400696003e-07,
+      "loss": 0.012,
+      "step": 4638
+    },
+    {
+      "epoch": 0.9214420498559937,
+      "grad_norm": 0.4766849044662014,
+      "learning_rate": 1.606566949282451e-07,
+      "loss": 0.0172,
+      "step": 4639
+    },
+    {
+      "epoch": 0.9216406793127421,
+      "grad_norm": 0.6710327474510234,
+      "learning_rate": 1.5984862981674786e-07,
+      "loss": 0.0129,
+      "step": 4640
+    },
+    {
+      "epoch": 0.9218393087694905,
+      "grad_norm": 0.42862810177240407,
+      "learning_rate": 1.5904256900708459e-07,
+      "loss": 0.0063,
+      "step": 4641
+    },
+    {
+      "epoch": 0.9220379382262389,
+      "grad_norm": 0.3757229399422139,
+      "learning_rate": 1.5823851283304546e-07,
+      "loss": 0.008,
+      "step": 4642
+    },
+    {
+      "epoch": 0.9222365676829873,
+      "grad_norm": 0.4649954354241579,
+      "learning_rate": 1.574364616275903e-07,
+      "loss": 0.009,
+      "step": 4643
+    },
+    {
+      "epoch": 0.9224351971397359,
+      "grad_norm": 0.24892525017019804,
+      "learning_rate": 1.5663641572284672e-07,
+      "loss": 0.0055,
+      "step": 4644
+    },
+    {
+      "epoch": 0.9226338265964843,
+      "grad_norm": 0.620937103802513,
+      "learning_rate": 1.5583837545011305e-07,
+      "loss": 0.0108,
+      "step": 4645
+    },
+    {
+      "epoch": 0.9228324560532327,
+      "grad_norm": 0.7464819124618941,
+      "learning_rate": 1.5504234113985661e-07,
+      "loss": 0.005,
+      "step": 4646
+    },
+    {
+      "epoch": 0.9230310855099811,
+      "grad_norm": 0.39950890377449805,
+      "learning_rate": 1.5424831312171595e-07,
+      "loss": 0.0091,
+      "step": 4647
+    },
+    {
+      "epoch": 0.9232297149667296,
+      "grad_norm": 0.7650531790637041,
+      "learning_rate": 1.5345629172449472e-07,
+      "loss": 0.0182,
+      "step": 4648
+    },
+    {
+      "epoch": 0.923428344423478,
+      "grad_norm": 0.4823999396347639,
+      "learning_rate": 1.5266627727617056e-07,
+      "loss": 0.0192,
+      "step": 4649
+    },
+    {
+      "epoch": 0.9236269738802264,
+      "grad_norm": 0.7401078670894918,
+      "learning_rate": 1.5187827010388677e-07,
+      "loss": 0.0127,
+      "step": 4650
+    },
+    {
+      "epoch": 0.9238256033369748,
+      "grad_norm": 0.3397450263276877,
+      "learning_rate": 1.510922705339557e-07,
+      "loss": 0.0108,
+      "step": 4651
+    },
+    {
+      "epoch": 0.9240242327937233,
+      "grad_norm": 0.3518773712824609,
+      "learning_rate": 1.503082788918603e-07,
+      "loss": 0.0057,
+      "step": 4652
+    },
+    {
+      "epoch": 0.9242228622504718,
+      "grad_norm": 0.2477018526451415,
+      "learning_rate": 1.4952629550224916e-07,
+      "loss": 0.0058,
+      "step": 4653
+    },
+    {
+      "epoch": 0.9244214917072202,
+      "grad_norm": 0.5414836310524654,
+      "learning_rate": 1.487463206889428e-07,
+      "loss": 0.0165,
+      "step": 4654
+    },
+    {
+      "epoch": 0.9246201211639686,
+      "grad_norm": 0.47446038228846893,
+      "learning_rate": 1.479683547749261e-07,
+      "loss": 0.0131,
+      "step": 4655
+    },
+    {
+      "epoch": 0.9248187506207171,
+      "grad_norm": 0.2636023459839493,
+      "learning_rate": 1.4719239808235418e-07,
+      "loss": 0.0041,
+      "step": 4656
+    },
+    {
+      "epoch": 0.9250173800774655,
+      "grad_norm": 0.40166636205399747,
+      "learning_rate": 1.464184509325517e-07,
+      "loss": 0.0055,
+      "step": 4657
+    },
+    {
+      "epoch": 0.9252160095342139,
+      "grad_norm": 0.44435793160055864,
+      "learning_rate": 1.4564651364600724e-07,
+      "loss": 0.0059,
+      "step": 4658
+    },
+    {
+      "epoch": 0.9254146389909623,
+      "grad_norm": 0.3994984612235425,
+      "learning_rate": 1.4487658654238123e-07,
+      "loss": 0.0071,
+      "step": 4659
+    },
+    {
+      "epoch": 0.9256132684477107,
+      "grad_norm": 0.47362741123170454,
+      "learning_rate": 1.4410866994049755e-07,
+      "loss": 0.011,
+      "step": 4660
+    },
+    {
+      "epoch": 0.9258118979044593,
+      "grad_norm": 0.7872100192173731,
+      "learning_rate": 1.433427641583518e-07,
+      "loss": 0.0169,
+      "step": 4661
+    },
+    {
+      "epoch": 0.9260105273612077,
+      "grad_norm": 0.8552459275140093,
+      "learning_rate": 1.4257886951310307e-07,
+      "loss": 0.0185,
+      "step": 4662
+    },
+    {
+      "epoch": 0.9262091568179561,
+      "grad_norm": 0.6608328624853623,
+      "learning_rate": 1.4181698632108055e-07,
+      "loss": 0.009,
+      "step": 4663
+    },
+    {
+      "epoch": 0.9264077862747045,
+      "grad_norm": 0.414776424248957,
+      "learning_rate": 1.4105711489777962e-07,
+      "loss": 0.0088,
+      "step": 4664
+    },
+    {
+      "epoch": 0.926606415731453,
+      "grad_norm": 0.6398136251284509,
+      "learning_rate": 1.4029925555786027e-07,
+      "loss": 0.0154,
+      "step": 4665
+    },
+    {
+      "epoch": 0.9268050451882014,
+      "grad_norm": 0.7474279649587607,
+      "learning_rate": 1.395434086151537e-07,
+      "loss": 0.0067,
+      "step": 4666
+    },
+    {
+      "epoch": 0.9270036746449498,
+      "grad_norm": 0.5064378151011119,
+      "learning_rate": 1.3878957438265338e-07,
+      "loss": 0.0116,
+      "step": 4667
+    },
+    {
+      "epoch": 0.9272023041016982,
+      "grad_norm": 0.45989881995150333,
+      "learning_rate": 1.3803775317252188e-07,
+      "loss": 0.0092,
+      "step": 4668
+    },
+    {
+      "epoch": 0.9274009335584468,
+      "grad_norm": 0.6599032509747758,
+      "learning_rate": 1.3728794529608846e-07,
+      "loss": 0.0093,
+      "step": 4669
+    },
+    {
+      "epoch": 0.9275995630151952,
+      "grad_norm": 0.5141005371853344,
+      "learning_rate": 1.3654015106384698e-07,
+      "loss": 0.0093,
+      "step": 4670
+    },
+    {
+      "epoch": 0.9277981924719436,
+      "grad_norm": 0.5941689240567093,
+      "learning_rate": 1.357943707854592e-07,
+      "loss": 0.0146,
+      "step": 4671
+    },
+    {
+      "epoch": 0.927996821928692,
+      "grad_norm": 0.20114286406354762,
+      "learning_rate": 1.3505060476975085e-07,
+      "loss": 0.003,
+      "step": 4672
+    },
+    {
+      "epoch": 0.9281954513854405,
+      "grad_norm": 0.45688487671874534,
+      "learning_rate": 1.3430885332471554e-07,
+      "loss": 0.0162,
+      "step": 4673
+    },
+    {
+      "epoch": 0.9283940808421889,
+      "grad_norm": 0.35614643694965575,
+      "learning_rate": 1.3356911675751093e-07,
+      "loss": 0.0051,
+      "step": 4674
+    },
+    {
+      "epoch": 0.9285927102989373,
+      "grad_norm": 0.5744415497903277,
+      "learning_rate": 1.3283139537446144e-07,
+      "loss": 0.0121,
+      "step": 4675
+    },
+    {
+      "epoch": 0.9287913397556857,
+      "grad_norm": 1.0850590038936332,
+      "learning_rate": 1.3209568948105768e-07,
+      "loss": 0.0145,
+      "step": 4676
+    },
+    {
+      "epoch": 0.9289899692124342,
+      "grad_norm": 0.45964179898690255,
+      "learning_rate": 1.313619993819537e-07,
+      "loss": 0.0131,
+      "step": 4677
+    },
+    {
+      "epoch": 0.9291885986691827,
+      "grad_norm": 0.5818030589778348,
+      "learning_rate": 1.3063032538097097e-07,
+      "loss": 0.0111,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9293872281259311,
+      "grad_norm": 0.6481989151858105,
+      "learning_rate": 1.2990066778109323e-07,
+      "loss": 0.0136,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9295858575826795,
+      "grad_norm": 0.37947609369661844,
+      "learning_rate": 1.2917302688447265e-07,
+      "loss": 0.0113,
+      "step": 4680
+    },
+    {
+      "epoch": 0.929784487039428,
+      "grad_norm": 0.5081483427859849,
+      "learning_rate": 1.284474029924232e-07,
+      "loss": 0.0103,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9299831164961764,
+      "grad_norm": 0.21444824245931055,
+      "learning_rate": 1.2772379640542564e-07,
+      "loss": 0.005,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9301817459529248,
+      "grad_norm": 0.26879234255685003,
+      "learning_rate": 1.270022074231253e-07,
+      "loss": 0.0099,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9303803754096732,
+      "grad_norm": 0.5160956745621564,
+      "learning_rate": 1.2628263634433035e-07,
+      "loss": 0.0138,
+      "step": 4684
+    },
+    {
+      "epoch": 0.9305790048664216,
+      "grad_norm": 0.49211468181460044,
+      "learning_rate": 1.2556508346701578e-07,
+      "loss": 0.0064,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9307776343231702,
+      "grad_norm": 0.8023564960097778,
+      "learning_rate": 1.2484954908831837e-07,
+      "loss": 0.0125,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9309762637799186,
+      "grad_norm": 0.9862722885789399,
+      "learning_rate": 1.241360335045405e-07,
+      "loss": 0.0085,
+      "step": 4687
+    },
+    {
+      "epoch": 0.931174893236667,
+      "grad_norm": 0.7887851867631835,
+      "learning_rate": 1.2342453701114864e-07,
+      "loss": 0.0131,
+      "step": 4688
+    },
+    {
+      "epoch": 0.9313735226934154,
+      "grad_norm": 0.3479006313592782,
+      "learning_rate": 1.227150599027721e-07,
+      "loss": 0.0094,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9315721521501639,
+      "grad_norm": 0.674927856372601,
+      "learning_rate": 1.2200760247320586e-07,
+      "loss": 0.0098,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9317707816069123,
+      "grad_norm": 0.42722316421638,
+      "learning_rate": 1.213021650154056e-07,
+      "loss": 0.0095,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9319694110636607,
+      "grad_norm": 0.301197655900226,
+      "learning_rate": 1.2059874782149317e-07,
+      "loss": 0.0061,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9321680405204091,
+      "grad_norm": 0.5133950310604148,
+      "learning_rate": 1.1989735118275337e-07,
+      "loss": 0.0127,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9323666699771576,
+      "grad_norm": 0.34014268074393605,
+      "learning_rate": 1.1919797538963274e-07,
+      "loss": 0.0059,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9325652994339061,
+      "grad_norm": 0.5164175565820704,
+      "learning_rate": 1.1850062073174351e-07,
+      "loss": 0.0116,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9327639288906545,
+      "grad_norm": 0.4070984340693311,
+      "learning_rate": 1.1780528749785802e-07,
+      "loss": 0.0077,
+      "step": 4696
+    },
+    {
+      "epoch": 0.9329625583474029,
+      "grad_norm": 0.38897032187757813,
+      "learning_rate": 1.1711197597591428e-07,
+      "loss": 0.0063,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9331611878041514,
+      "grad_norm": 0.2930844011928402,
+      "learning_rate": 1.1642068645301152e-07,
+      "loss": 0.0053,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9333598172608998,
+      "grad_norm": 0.3764839957960558,
+      "learning_rate": 1.1573141921541131e-07,
+      "loss": 0.0054,
+      "step": 4699
+    },
+    {
+      "epoch": 0.9335584467176482,
+      "grad_norm": 0.43523780379200877,
+      "learning_rate": 1.1504417454854033e-07,
+      "loss": 0.013,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9337570761743966,
+      "grad_norm": 0.5750229920284355,
+      "learning_rate": 1.1435895273698372e-07,
+      "loss": 0.0084,
+      "step": 4701
+    },
+    {
+      "epoch": 0.933955705631145,
+      "grad_norm": 0.19886700519682282,
+      "learning_rate": 1.1367575406449282e-07,
+      "loss": 0.0046,
+      "step": 4702
+    },
+    {
+      "epoch": 0.9341543350878936,
+      "grad_norm": 0.790833202260621,
+      "learning_rate": 1.1299457881397858e-07,
+      "loss": 0.0185,
+      "step": 4703
+    },
+    {
+      "epoch": 0.934352964544642,
+      "grad_norm": 0.4219918573113129,
+      "learning_rate": 1.1231542726751532e-07,
+      "loss": 0.0092,
+      "step": 4704
+    },
+    {
+      "epoch": 0.9345515940013904,
+      "grad_norm": 0.7230909029282245,
+      "learning_rate": 1.1163829970633865e-07,
+      "loss": 0.0147,
+      "step": 4705
+    },
+    {
+      "epoch": 0.9347502234581389,
+      "grad_norm": 0.5093534424476391,
+      "learning_rate": 1.1096319641084708e-07,
+      "loss": 0.0116,
+      "step": 4706
+    },
+    {
+      "epoch": 0.9349488529148873,
+      "grad_norm": 0.27347946955908975,
+      "learning_rate": 1.1029011766059972e-07,
+      "loss": 0.0034,
+      "step": 4707
+    },
+    {
+      "epoch": 0.9351474823716357,
+      "grad_norm": 1.0701505055181986,
+      "learning_rate": 1.0961906373431808e-07,
+      "loss": 0.0117,
+      "step": 4708
+    },
+    {
+      "epoch": 0.9353461118283841,
+      "grad_norm": 0.6516188206878525,
+      "learning_rate": 1.0895003490988487e-07,
+      "loss": 0.0114,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9355447412851325,
+      "grad_norm": 0.4902354426875574,
+      "learning_rate": 1.0828303146434404e-07,
+      "loss": 0.0088,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9357433707418811,
+      "grad_norm": 0.5209392249997609,
+      "learning_rate": 1.0761805367390187e-07,
+      "loss": 0.0117,
+      "step": 4711
+    },
+    {
+      "epoch": 0.9359420001986295,
+      "grad_norm": 1.259910339066219,
+      "learning_rate": 1.0695510181392365e-07,
+      "loss": 0.0088,
+      "step": 4712
+    },
+    {
+      "epoch": 0.9361406296553779,
+      "grad_norm": 0.5151392556944235,
+      "learning_rate": 1.0629417615893756e-07,
+      "loss": 0.0139,
+      "step": 4713
+    },
+    {
+      "epoch": 0.9363392591121263,
+      "grad_norm": 0.9491713148601669,
+      "learning_rate": 1.0563527698263298e-07,
+      "loss": 0.018,
+      "step": 4714
+    },
+    {
+      "epoch": 0.9365378885688748,
+      "grad_norm": 1.2153903329483169,
+      "learning_rate": 1.0497840455785835e-07,
+      "loss": 0.0173,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9367365180256232,
+      "grad_norm": 0.288079961745967,
+      "learning_rate": 1.0432355915662496e-07,
+      "loss": 0.0033,
+      "step": 4716
+    },
+    {
+      "epoch": 0.9369351474823716,
+      "grad_norm": 0.3536849277399397,
+      "learning_rate": 1.0367074105010256e-07,
+      "loss": 0.0075,
+      "step": 4717
+    },
+    {
+      "epoch": 0.93713377693912,
+      "grad_norm": 0.6473193451222942,
+      "learning_rate": 1.0301995050862323e-07,
+      "loss": 0.0126,
+      "step": 4718
+    },
+    {
+      "epoch": 0.9373324063958685,
+      "grad_norm": 0.6978153920712078,
+      "learning_rate": 1.0237118780167809e-07,
+      "loss": 0.0106,
+      "step": 4719
+    },
+    {
+      "epoch": 0.937531035852617,
+      "grad_norm": 0.41837417287713097,
+      "learning_rate": 1.0172445319792002e-07,
+      "loss": 0.009,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9377296653093654,
+      "grad_norm": 0.3535301700831529,
+      "learning_rate": 1.0107974696516032e-07,
+      "loss": 0.0078,
+      "step": 4721
+    },
+    {
+      "epoch": 0.9379282947661138,
+      "grad_norm": 0.7905465076037768,
+      "learning_rate": 1.0043706937037156e-07,
+      "loss": 0.0134,
+      "step": 4722
+    },
+    {
+      "epoch": 0.9381269242228623,
+      "grad_norm": 0.36966492424476793,
+      "learning_rate": 9.979642067968587e-08,
+      "loss": 0.0116,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9383255536796107,
+      "grad_norm": 0.6252230556557902,
+      "learning_rate": 9.915780115839491e-08,
+      "loss": 0.0116,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9385241831363591,
+      "grad_norm": 0.8581563383955567,
+      "learning_rate": 9.852121107095047e-08,
+      "loss": 0.0266,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9387228125931075,
+      "grad_norm": 0.5722638109056252,
+      "learning_rate": 9.788665068096504e-08,
+      "loss": 0.0162,
+      "step": 4726
+    },
+    {
+      "epoch": 0.938921442049856,
+      "grad_norm": 0.7084688510586287,
+      "learning_rate": 9.725412025120783e-08,
+      "loss": 0.0087,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9391200715066045,
+      "grad_norm": 0.1863360147125236,
+      "learning_rate": 9.662362004360992e-08,
+      "loss": 0.0036,
+      "step": 4728
+    },
+    {
+      "epoch": 0.9393187009633529,
+      "grad_norm": 0.5073294558989344,
+      "learning_rate": 9.599515031926021e-08,
+      "loss": 0.0093,
+      "step": 4729
+    },
+    {
+      "epoch": 0.9395173304201013,
+      "grad_norm": 0.8287013770837947,
+      "learning_rate": 9.536871133840775e-08,
+      "loss": 0.0186,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9397159598768497,
+      "grad_norm": 0.4145640260838776,
+      "learning_rate": 9.474430336046059e-08,
+      "loss": 0.011,
+      "step": 4731
+    },
+    {
+      "epoch": 0.9399145893335982,
+      "grad_norm": 0.3121093385078184,
+      "learning_rate": 9.412192664398467e-08,
+      "loss": 0.004,
+      "step": 4732
+    },
+    {
+      "epoch": 0.9401132187903466,
+      "grad_norm": 0.5686734537347415,
+      "learning_rate": 9.350158144670662e-08,
+      "loss": 0.0153,
+      "step": 4733
+    },
+    {
+      "epoch": 0.940311848247095,
+      "grad_norm": 1.060966428278419,
+      "learning_rate": 9.28832680255104e-08,
+      "loss": 0.019,
+      "step": 4734
+    },
+    {
+      "epoch": 0.9405104777038434,
+      "grad_norm": 0.4224352499347425,
+      "learning_rate": 9.22669866364384e-08,
+      "loss": 0.0067,
+      "step": 4735
+    },
+    {
+      "epoch": 0.9407091071605919,
+      "grad_norm": 0.3341915120102901,
+      "learning_rate": 9.165273753469261e-08,
+      "loss": 0.0069,
+      "step": 4736
+    },
+    {
+      "epoch": 0.9409077366173404,
+      "grad_norm": 0.6457098860974007,
+      "learning_rate": 9.10405209746329e-08,
+      "loss": 0.0137,
+      "step": 4737
+    },
+    {
+      "epoch": 0.9411063660740888,
+      "grad_norm": 0.24789794897733472,
+      "learning_rate": 9.043033720977756e-08,
+      "loss": 0.0048,
+      "step": 4738
+    },
+    {
+      "epoch": 0.9413049955308372,
+      "grad_norm": 0.3608454499887148,
+      "learning_rate": 8.982218649280284e-08,
+      "loss": 0.0122,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9415036249875857,
+      "grad_norm": 0.7509200650856567,
+      "learning_rate": 8.921606907554337e-08,
+      "loss": 0.0116,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9417022544443341,
+      "grad_norm": 0.3689801779957159,
+      "learning_rate": 8.861198520899172e-08,
+      "loss": 0.0076,
+      "step": 4741
+    },
+    {
+      "epoch": 0.9419008839010825,
+      "grad_norm": 0.5557873310711061,
+      "learning_rate": 8.800993514329892e-08,
+      "loss": 0.0153,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9420995133578309,
+      "grad_norm": 0.37590619822900495,
+      "learning_rate": 8.74099191277733e-08,
+      "loss": 0.0091,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9422981428145794,
+      "grad_norm": 0.8370582813494728,
+      "learning_rate": 8.681193741088e-08,
+      "loss": 0.0087,
+      "step": 4744
+    },
+    {
+      "epoch": 0.9424967722713279,
+      "grad_norm": 0.9398151319661212,
+      "learning_rate": 8.621599024024374e-08,
+      "loss": 0.0148,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9426954017280763,
+      "grad_norm": 0.5122269675023894,
+      "learning_rate": 8.562207786264487e-08,
+      "loss": 0.0089,
+      "step": 4746
+    },
+    {
+      "epoch": 0.9428940311848247,
+      "grad_norm": 0.6367900598425142,
+      "learning_rate": 8.503020052402223e-08,
+      "loss": 0.0109,
+      "step": 4747
+    },
+    {
+      "epoch": 0.9430926606415732,
+      "grad_norm": 0.7304382048017897,
+      "learning_rate": 8.444035846947141e-08,
+      "loss": 0.0169,
+      "step": 4748
+    },
+    {
+      "epoch": 0.9432912900983216,
+      "grad_norm": 0.7902391038165741,
+      "learning_rate": 8.385255194324593e-08,
+      "loss": 0.0119,
+      "step": 4749
+    },
+    {
+      "epoch": 0.94348991955507,
+      "grad_norm": 0.403170445331138,
+      "learning_rate": 8.326678118875554e-08,
+      "loss": 0.0101,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9436885490118184,
+      "grad_norm": 0.4907736576120689,
+      "learning_rate": 8.268304644856673e-08,
+      "loss": 0.0108,
+      "step": 4751
+    },
+    {
+      "epoch": 0.9438871784685668,
+      "grad_norm": 0.5505554874481177,
+      "learning_rate": 8.210134796440449e-08,
+      "loss": 0.0107,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9440858079253154,
+      "grad_norm": 0.5428734384940673,
+      "learning_rate": 8.152168597714894e-08,
+      "loss": 0.0093,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9442844373820638,
+      "grad_norm": 0.44060876168794505,
+      "learning_rate": 8.094406072683858e-08,
+      "loss": 0.0062,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9444830668388122,
+      "grad_norm": 0.40725989340402463,
+      "learning_rate": 8.036847245266543e-08,
+      "loss": 0.0099,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9446816962955606,
+      "grad_norm": 0.46261890604962713,
+      "learning_rate": 7.979492139298162e-08,
+      "loss": 0.011,
+      "step": 4756
+    },
+    {
+      "epoch": 0.9448803257523091,
+      "grad_norm": 0.4976138795208467,
+      "learning_rate": 7.92234077852938e-08,
+      "loss": 0.0136,
+      "step": 4757
+    },
+    {
+      "epoch": 0.9450789552090575,
+      "grad_norm": 0.5403839208497061,
+      "learning_rate": 7.865393186626491e-08,
+      "loss": 0.0068,
+      "step": 4758
+    },
+    {
+      "epoch": 0.9452775846658059,
+      "grad_norm": 0.34280255446442026,
+      "learning_rate": 7.808649387171519e-08,
+      "loss": 0.0063,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9454762141225543,
+      "grad_norm": 0.9007767217887392,
+      "learning_rate": 7.752109403661834e-08,
+      "loss": 0.011,
+      "step": 4760
+    },
+    {
+      "epoch": 0.9456748435793028,
+      "grad_norm": 1.0223302663255138,
+      "learning_rate": 7.695773259510764e-08,
+      "loss": 0.0094,
+      "step": 4761
+    },
+    {
+      "epoch": 0.9458734730360513,
+      "grad_norm": 0.41230357058037975,
+      "learning_rate": 7.639640978046981e-08,
+      "loss": 0.0076,
+      "step": 4762
+    },
+    {
+      "epoch": 0.9460721024927997,
+      "grad_norm": 0.4318170038225908,
+      "learning_rate": 7.583712582514724e-08,
+      "loss": 0.0121,
+      "step": 4763
+    },
+    {
+      "epoch": 0.9462707319495481,
+      "grad_norm": 0.2930715164840165,
+      "learning_rate": 7.527988096074079e-08,
+      "loss": 0.004,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9464693614062966,
+      "grad_norm": 1.1647632495336515,
+      "learning_rate": 7.47246754180031e-08,
+      "loss": 0.0114,
+      "step": 4765
+    },
+    {
+      "epoch": 0.946667990863045,
+      "grad_norm": 0.5002685742542026,
+      "learning_rate": 7.417150942684525e-08,
+      "loss": 0.0115,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9468666203197934,
+      "grad_norm": 0.5858011863544192,
+      "learning_rate": 7.362038321633235e-08,
+      "loss": 0.0086,
+      "step": 4767
+    },
+    {
+      "epoch": 0.9470652497765418,
+      "grad_norm": 0.5066199298525029,
+      "learning_rate": 7.307129701468574e-08,
+      "loss": 0.0103,
+      "step": 4768
+    },
+    {
+      "epoch": 0.9472638792332903,
+      "grad_norm": 0.2689653914202334,
+      "learning_rate": 7.252425104928074e-08,
+      "loss": 0.0064,
+      "step": 4769
+    },
+    {
+      "epoch": 0.9474625086900388,
+      "grad_norm": 0.6049148054355731,
+      "learning_rate": 7.197924554664893e-08,
+      "loss": 0.0119,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9476611381467872,
+      "grad_norm": 0.5517727484698056,
+      "learning_rate": 7.1436280732477e-08,
+      "loss": 0.0132,
+      "step": 4771
+    },
+    {
+      "epoch": 0.9478597676035356,
+      "grad_norm": 0.32666164342892834,
+      "learning_rate": 7.089535683160508e-08,
+      "loss": 0.0062,
+      "step": 4772
+    },
+    {
+      "epoch": 0.948058397060284,
+      "grad_norm": 1.102212142762428,
+      "learning_rate": 7.035647406803015e-08,
+      "loss": 0.0136,
+      "step": 4773
+    },
+    {
+      "epoch": 0.9482570265170325,
+      "grad_norm": 0.7658504815210466,
+      "learning_rate": 6.981963266490199e-08,
+      "loss": 0.0116,
+      "step": 4774
+    },
+    {
+      "epoch": 0.9484556559737809,
+      "grad_norm": 0.8130273192518539,
+      "learning_rate": 6.92848328445267e-08,
+      "loss": 0.0128,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9486542854305293,
+      "grad_norm": 0.6824866401998253,
+      "learning_rate": 6.875207482836544e-08,
+      "loss": 0.0164,
+      "step": 4776
+    },
+    {
+      "epoch": 0.9488529148872777,
+      "grad_norm": 0.47569708326658633,
+      "learning_rate": 6.822135883703063e-08,
+      "loss": 0.0089,
+      "step": 4777
+    },
+    {
+      "epoch": 0.9490515443440262,
+      "grad_norm": 0.38871295881882434,
+      "learning_rate": 6.769268509029315e-08,
+      "loss": 0.0124,
+      "step": 4778
+    },
+    {
+      "epoch": 0.9492501738007747,
+      "grad_norm": 0.6378821616400225,
+      "learning_rate": 6.716605380707508e-08,
+      "loss": 0.0104,
+      "step": 4779
+    },
+    {
+      "epoch": 0.9494488032575231,
+      "grad_norm": 0.745002833768245,
+      "learning_rate": 6.664146520545422e-08,
+      "loss": 0.0115,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9496474327142715,
+      "grad_norm": 0.3103526017311052,
+      "learning_rate": 6.611891950266235e-08,
+      "loss": 0.006,
+      "step": 4781
+    },
+    {
+      "epoch": 0.94984606217102,
+      "grad_norm": 0.44138783004158555,
+      "learning_rate": 6.559841691508473e-08,
+      "loss": 0.0109,
+      "step": 4782
+    },
+    {
+      "epoch": 0.9500446916277684,
+      "grad_norm": 0.6016533565370209,
+      "learning_rate": 6.507995765826169e-08,
+      "loss": 0.0135,
+      "step": 4783
+    },
+    {
+      "epoch": 0.9502433210845168,
+      "grad_norm": 0.9664137702511328,
+      "learning_rate": 6.456354194688597e-08,
+      "loss": 0.0103,
+      "step": 4784
+    },
+    {
+      "epoch": 0.9504419505412652,
+      "grad_norm": 0.2967531583933644,
+      "learning_rate": 6.404916999480482e-08,
+      "loss": 0.0061,
+      "step": 4785
+    },
+    {
+      "epoch": 0.9506405799980137,
+      "grad_norm": 0.4734645839133741,
+      "learning_rate": 6.353684201502008e-08,
+      "loss": 0.0094,
+      "step": 4786
+    },
+    {
+      "epoch": 0.9508392094547622,
+      "grad_norm": 0.7133914033624615,
+      "learning_rate": 6.302655821968485e-08,
+      "loss": 0.0146,
+      "step": 4787
+    },
+    {
+      "epoch": 0.9510378389115106,
+      "grad_norm": 0.4600218420642334,
+      "learning_rate": 6.25183188201084e-08,
+      "loss": 0.0097,
+      "step": 4788
+    },
+    {
+      "epoch": 0.951236468368259,
+      "grad_norm": 0.38688423685277884,
+      "learning_rate": 6.201212402675072e-08,
+      "loss": 0.0084,
+      "step": 4789
+    },
+    {
+      "epoch": 0.9514350978250075,
+      "grad_norm": 0.566156909391181,
+      "learning_rate": 6.15079740492286e-08,
+      "loss": 0.0127,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9516337272817559,
+      "grad_norm": 0.5610189586192836,
+      "learning_rate": 6.100586909630779e-08,
+      "loss": 0.0126,
+      "step": 4791
+    },
+    {
+      "epoch": 0.9518323567385043,
+      "grad_norm": 0.6857347036976907,
+      "learning_rate": 6.050580937591144e-08,
+      "loss": 0.0117,
+      "step": 4792
+    },
+    {
+      "epoch": 0.9520309861952527,
+      "grad_norm": 0.5458484629152233,
+      "learning_rate": 6.000779509511279e-08,
+      "loss": 0.0115,
+      "step": 4793
+    },
+    {
+      "epoch": 0.9522296156520011,
+      "grad_norm": 0.6424256763252679,
+      "learning_rate": 5.951182646013853e-08,
+      "loss": 0.0161,
+      "step": 4794
+    },
+    {
+      "epoch": 0.9524282451087497,
+      "grad_norm": 0.544918430900207,
+      "learning_rate": 5.901790367636995e-08,
+      "loss": 0.0138,
+      "step": 4795
+    },
+    {
+      "epoch": 0.9526268745654981,
+      "grad_norm": 0.7253601543796748,
+      "learning_rate": 5.8526026948338974e-08,
+      "loss": 0.0188,
+      "step": 4796
+    },
+    {
+      "epoch": 0.9528255040222465,
+      "grad_norm": 0.6596213294085469,
+      "learning_rate": 5.803619647973213e-08,
+      "loss": 0.0108,
+      "step": 4797
+    },
+    {
+      "epoch": 0.953024133478995,
+      "grad_norm": 0.3094292663332667,
+      "learning_rate": 5.754841247338716e-08,
+      "loss": 0.0075,
+      "step": 4798
+    },
+    {
+      "epoch": 0.9532227629357434,
+      "grad_norm": 0.2653176462450698,
+      "learning_rate": 5.706267513129527e-08,
+      "loss": 0.006,
+      "step": 4799
+    },
+    {
+      "epoch": 0.9534213923924918,
+      "grad_norm": 0.2862839957357303,
+      "learning_rate": 5.657898465459943e-08,
+      "loss": 0.007,
+      "step": 4800
+    },
+    {
+      "epoch": 0.9536200218492402,
+      "grad_norm": 0.46267765672416544,
+      "learning_rate": 5.609734124359556e-08,
+      "loss": 0.0115,
+      "step": 4801
+    },
+    {
+      "epoch": 0.9538186513059886,
+      "grad_norm": 1.3552174604785634,
+      "learning_rate": 5.5617745097731876e-08,
+      "loss": 0.0066,
+      "step": 4802
+    },
+    {
+      "epoch": 0.9540172807627371,
+      "grad_norm": 0.6673917782595586,
+      "learning_rate": 5.5140196415608414e-08,
+      "loss": 0.0128,
+      "step": 4803
+    },
+    {
+      "epoch": 0.9542159102194856,
+      "grad_norm": 0.6949616822551641,
+      "learning_rate": 5.466469539497809e-08,
+      "loss": 0.0158,
+      "step": 4804
+    },
+    {
+      "epoch": 0.954414539676234,
+      "grad_norm": 0.7623499887827077,
+      "learning_rate": 5.419124223274452e-08,
+      "loss": 0.0139,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9546131691329824,
+      "grad_norm": 0.548628908422014,
+      "learning_rate": 5.371983712496476e-08,
+      "loss": 0.0122,
+      "step": 4806
+    },
+    {
+      "epoch": 0.9548117985897309,
+      "grad_norm": 0.8368126791005601,
+      "learning_rate": 5.325048026684765e-08,
+      "loss": 0.0156,
+      "step": 4807
+    },
+    {
+      "epoch": 0.9550104280464793,
+      "grad_norm": 0.5712358763963159,
+      "learning_rate": 5.278317185275217e-08,
+      "loss": 0.0143,
+      "step": 4808
+    },
+    {
+      "epoch": 0.9552090575032277,
+      "grad_norm": 0.33369125553839335,
+      "learning_rate": 5.23179120761913e-08,
+      "loss": 0.0091,
+      "step": 4809
+    },
+    {
+      "epoch": 0.9554076869599761,
+      "grad_norm": 0.5426548631897875,
+      "learning_rate": 5.185470112982816e-08,
+      "loss": 0.01,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9556063164167246,
+      "grad_norm": 0.7736330728886897,
+      "learning_rate": 5.139353920547818e-08,
+      "loss": 0.0084,
+      "step": 4811
+    },
+    {
+      "epoch": 0.9558049458734731,
+      "grad_norm": 0.7726944536488665,
+      "learning_rate": 5.093442649410807e-08,
+      "loss": 0.0122,
+      "step": 4812
+    },
+    {
+      "epoch": 0.9560035753302215,
+      "grad_norm": 0.41025603735219146,
+      "learning_rate": 5.0477363185835736e-08,
+      "loss": 0.0082,
+      "step": 4813
+    },
+    {
+      "epoch": 0.9562022047869699,
+      "grad_norm": 0.42810031719834896,
+      "learning_rate": 5.0022349469930344e-08,
+      "loss": 0.0076,
+      "step": 4814
+    },
+    {
+      "epoch": 0.9564008342437184,
+      "grad_norm": 0.5312298899611213,
+      "learning_rate": 4.9569385534813386e-08,
+      "loss": 0.0091,
+      "step": 4815
+    },
+    {
+      "epoch": 0.9565994637004668,
+      "grad_norm": 0.5843424770418346,
+      "learning_rate": 4.911847156805649e-08,
+      "loss": 0.0058,
+      "step": 4816
+    },
+    {
+      "epoch": 0.9567980931572152,
+      "grad_norm": 0.627548785238714,
+      "learning_rate": 4.866960775638252e-08,
+      "loss": 0.0131,
+      "step": 4817
+    },
+    {
+      "epoch": 0.9569967226139636,
+      "grad_norm": 0.7105539318515306,
+      "learning_rate": 4.8222794285665006e-08,
+      "loss": 0.0183,
+      "step": 4818
+    },
+    {
+      "epoch": 0.957195352070712,
+      "grad_norm": 0.6002583148423645,
+      "learning_rate": 4.7778031340930397e-08,
+      "loss": 0.0107,
+      "step": 4819
+    },
+    {
+      "epoch": 0.9573939815274605,
+      "grad_norm": 0.6818174820977825,
+      "learning_rate": 4.7335319106353026e-08,
+      "loss": 0.0168,
+      "step": 4820
+    },
+    {
+      "epoch": 0.957592610984209,
+      "grad_norm": 0.3965043691901418,
+      "learning_rate": 4.689465776526125e-08,
+      "loss": 0.0083,
+      "step": 4821
+    },
+    {
+      "epoch": 0.9577912404409574,
+      "grad_norm": 0.6228622351159547,
+      "learning_rate": 4.645604750013078e-08,
+      "loss": 0.009,
+      "step": 4822
+    },
+    {
+      "epoch": 0.9579898698977058,
+      "grad_norm": 0.47410239295040185,
+      "learning_rate": 4.601948849259019e-08,
+      "loss": 0.014,
+      "step": 4823
+    },
+    {
+      "epoch": 0.9581884993544543,
+      "grad_norm": 1.0847330066440664,
+      "learning_rate": 4.558498092341879e-08,
+      "loss": 0.016,
+      "step": 4824
+    },
+    {
+      "epoch": 0.9583871288112027,
+      "grad_norm": 1.061431136868962,
+      "learning_rate": 4.515252497254541e-08,
+      "loss": 0.0147,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9585857582679511,
+      "grad_norm": 0.4114241358058321,
+      "learning_rate": 4.4722120819049586e-08,
+      "loss": 0.0111,
+      "step": 4826
+    },
+    {
+      "epoch": 0.9587843877246995,
+      "grad_norm": 0.6247463058011158,
+      "learning_rate": 4.4293768641160416e-08,
+      "loss": 0.0099,
+      "step": 4827
+    },
+    {
+      "epoch": 0.958983017181448,
+      "grad_norm": 0.8021242243802347,
+      "learning_rate": 4.38674686162599e-08,
+      "loss": 0.0136,
+      "step": 4828
+    },
+    {
+      "epoch": 0.9591816466381965,
+      "grad_norm": 0.5276059971971032,
+      "learning_rate": 4.344322092087683e-08,
+      "loss": 0.0139,
+      "step": 4829
+    },
+    {
+      "epoch": 0.9593802760949449,
+      "grad_norm": 0.4836014882831527,
+      "learning_rate": 4.302102573069289e-08,
+      "loss": 0.0115,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9595789055516933,
+      "grad_norm": 0.34443701008701366,
+      "learning_rate": 4.260088322053768e-08,
+      "loss": 0.0123,
+      "step": 4831
+    },
+    {
+      "epoch": 0.9597775350084418,
+      "grad_norm": 0.3859441891012454,
+      "learning_rate": 4.2182793564392034e-08,
+      "loss": 0.0092,
+      "step": 4832
+    },
+    {
+      "epoch": 0.9599761644651902,
+      "grad_norm": 0.5084520994177352,
+      "learning_rate": 4.176675693538745e-08,
+      "loss": 0.013,
+      "step": 4833
+    },
+    {
+      "epoch": 0.9601747939219386,
+      "grad_norm": 0.3918431048956989,
+      "learning_rate": 4.13527735058028e-08,
+      "loss": 0.0069,
+      "step": 4834
+    },
+    {
+      "epoch": 0.960373423378687,
+      "grad_norm": 0.5475933146987769,
+      "learning_rate": 4.094084344706928e-08,
+      "loss": 0.0131,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9605720528354355,
+      "grad_norm": 0.8255199408065383,
+      "learning_rate": 4.053096692976655e-08,
+      "loss": 0.0127,
+      "step": 4836
+    },
+    {
+      "epoch": 0.960770682292184,
+      "grad_norm": 0.9276148823282381,
+      "learning_rate": 4.012314412362328e-08,
+      "loss": 0.0119,
+      "step": 4837
+    },
+    {
+      "epoch": 0.9609693117489324,
+      "grad_norm": 0.5402494802195912,
+      "learning_rate": 3.971737519751939e-08,
+      "loss": 0.0083,
+      "step": 4838
+    },
+    {
+      "epoch": 0.9611679412056808,
+      "grad_norm": 0.35876195272358025,
+      "learning_rate": 3.9313660319483246e-08,
+      "loss": 0.0085,
+      "step": 4839
+    },
+    {
+      "epoch": 0.9613665706624293,
+      "grad_norm": 0.42863726316703926,
+      "learning_rate": 3.8911999656692787e-08,
+      "loss": 0.0053,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9615652001191777,
+      "grad_norm": 0.43862831859121126,
+      "learning_rate": 3.851239337547441e-08,
+      "loss": 0.0099,
+      "step": 4841
+    },
+    {
+      "epoch": 0.9617638295759261,
+      "grad_norm": 0.5380419062452532,
+      "learning_rate": 3.8114841641305744e-08,
+      "loss": 0.0102,
+      "step": 4842
+    },
+    {
+      "epoch": 0.9619624590326745,
+      "grad_norm": 0.3862160575949929,
+      "learning_rate": 3.7719344618812326e-08,
+      "loss": 0.0121,
+      "step": 4843
+    },
+    {
+      "epoch": 0.9621610884894229,
+      "grad_norm": 0.4397239298419148,
+      "learning_rate": 3.7325902471768706e-08,
+      "loss": 0.0095,
+      "step": 4844
+    },
+    {
+      "epoch": 0.9623597179461714,
+      "grad_norm": 0.8870761334367195,
+      "learning_rate": 3.693451536309955e-08,
+      "loss": 0.018,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9625583474029199,
+      "grad_norm": 0.7392267280358624,
+      "learning_rate": 3.65451834548769e-08,
+      "loss": 0.0132,
+      "step": 4846
+    },
+    {
+      "epoch": 0.9627569768596683,
+      "grad_norm": 0.9086319168545279,
+      "learning_rate": 3.6157906908323995e-08,
+      "loss": 0.0112,
+      "step": 4847
+    },
+    {
+      "epoch": 0.9629556063164167,
+      "grad_norm": 0.5583448293826135,
+      "learning_rate": 3.5772685883809775e-08,
+      "loss": 0.008,
+      "step": 4848
+    },
+    {
+      "epoch": 0.9631542357731652,
+      "grad_norm": 0.5863970387552007,
+      "learning_rate": 3.5389520540856094e-08,
+      "loss": 0.0062,
+      "step": 4849
+    },
+    {
+      "epoch": 0.9633528652299136,
+      "grad_norm": 0.25369702479905915,
+      "learning_rate": 3.500841103812991e-08,
+      "loss": 0.006,
+      "step": 4850
+    },
+    {
+      "epoch": 0.963551494686662,
+      "grad_norm": 0.7699812014005178,
+      "learning_rate": 3.462935753344832e-08,
+      "loss": 0.0124,
+      "step": 4851
+    },
+    {
+      "epoch": 0.9637501241434104,
+      "grad_norm": 0.39448570416002493,
+      "learning_rate": 3.4252360183777976e-08,
+      "loss": 0.0068,
+      "step": 4852
+    },
+    {
+      "epoch": 0.9639487536001589,
+      "grad_norm": 0.37419322323165516,
+      "learning_rate": 3.38774191452318e-08,
+      "loss": 0.0063,
+      "step": 4853
+    },
+    {
+      "epoch": 0.9641473830569074,
+      "grad_norm": 0.6355343515700734,
+      "learning_rate": 3.350453457307335e-08,
+      "loss": 0.0141,
+      "step": 4854
+    },
+    {
+      "epoch": 0.9643460125136558,
+      "grad_norm": 0.8356043327709957,
+      "learning_rate": 3.313370662171411e-08,
+      "loss": 0.0197,
+      "step": 4855
+    },
+    {
+      "epoch": 0.9645446419704042,
+      "grad_norm": 0.88373860293526,
+      "learning_rate": 3.276493544471237e-08,
+      "loss": 0.0137,
+      "step": 4856
+    },
+    {
+      "epoch": 0.9647432714271527,
+      "grad_norm": 0.43188108347987786,
+      "learning_rate": 3.239822119477709e-08,
+      "loss": 0.0099,
+      "step": 4857
+    },
+    {
+      "epoch": 0.9649419008839011,
+      "grad_norm": 0.2922777433644811,
+      "learning_rate": 3.2033564023762895e-08,
+      "loss": 0.0076,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9651405303406495,
+      "grad_norm": 0.41982816990822347,
+      "learning_rate": 3.167096408267567e-08,
+      "loss": 0.0064,
+      "step": 4859
+    },
+    {
+      "epoch": 0.9653391597973979,
+      "grad_norm": 0.3912328421574355,
+      "learning_rate": 3.131042152166641e-08,
+      "loss": 0.0078,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9655377892541464,
+      "grad_norm": 0.49125451640633777,
+      "learning_rate": 3.0951936490035696e-08,
+      "loss": 0.0071,
+      "step": 4861
+    },
+    {
+      "epoch": 0.9657364187108948,
+      "grad_norm": 0.38541352755974334,
+      "learning_rate": 3.059550913623199e-08,
+      "loss": 0.0106,
+      "step": 4862
+    },
+    {
+      "epoch": 0.9659350481676433,
+      "grad_norm": 0.34748274835246395,
+      "learning_rate": 3.024113960785169e-08,
+      "loss": 0.0068,
+      "step": 4863
+    },
+    {
+      "epoch": 0.9661336776243917,
+      "grad_norm": 0.33066145540841996,
+      "learning_rate": 2.9888828051638505e-08,
+      "loss": 0.0088,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9663323070811402,
+      "grad_norm": 0.21097836166400444,
+      "learning_rate": 2.9538574613484084e-08,
+      "loss": 0.0058,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9665309365378886,
+      "grad_norm": 0.36457461980716876,
+      "learning_rate": 2.919037943842906e-08,
+      "loss": 0.0062,
+      "step": 4866
+    },
+    {
+      "epoch": 0.966729565994637,
+      "grad_norm": 0.42668003224731427,
+      "learning_rate": 2.884424267065922e-08,
+      "loss": 0.0083,
+      "step": 4867
+    },
+    {
+      "epoch": 0.9669281954513854,
+      "grad_norm": 0.8275980493196213,
+      "learning_rate": 2.8500164453511002e-08,
+      "loss": 0.019,
+      "step": 4868
+    },
+    {
+      "epoch": 0.9671268249081338,
+      "grad_norm": 0.3290524233842703,
+      "learning_rate": 2.8158144929466e-08,
+      "loss": 0.0066,
+      "step": 4869
+    },
+    {
+      "epoch": 0.9673254543648823,
+      "grad_norm": 0.372833561243726,
+      "learning_rate": 2.78181842401537e-08,
+      "loss": 0.0045,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9675240838216308,
+      "grad_norm": 0.4524320356338434,
+      "learning_rate": 2.74802825263526e-08,
+      "loss": 0.0061,
+      "step": 4871
+    },
+    {
+      "epoch": 0.9677227132783792,
+      "grad_norm": 0.5396272827787818,
+      "learning_rate": 2.714443992798632e-08,
+      "loss": 0.0114,
+      "step": 4872
+    },
+    {
+      "epoch": 0.9679213427351276,
+      "grad_norm": 0.71897085626895,
+      "learning_rate": 2.681065658412807e-08,
+      "loss": 0.0149,
+      "step": 4873
+    },
+    {
+      "epoch": 0.9681199721918761,
+      "grad_norm": 0.3973267868046733,
+      "learning_rate": 2.6478932632996724e-08,
+      "loss": 0.0071,
+      "step": 4874
+    },
+    {
+      "epoch": 0.9683186016486245,
+      "grad_norm": 0.7077697157290327,
+      "learning_rate": 2.6149268211957955e-08,
+      "loss": 0.0185,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9685172311053729,
+      "grad_norm": 0.5366383773695785,
+      "learning_rate": 2.5821663457527013e-08,
+      "loss": 0.0115,
+      "step": 4876
+    },
+    {
+      "epoch": 0.9687158605621213,
+      "grad_norm": 0.587638684038049,
+      "learning_rate": 2.549611850536371e-08,
+      "loss": 0.0098,
+      "step": 4877
+    },
+    {
+      "epoch": 0.9689144900188698,
+      "grad_norm": 0.5064559833862212,
+      "learning_rate": 2.517263349027632e-08,
+      "loss": 0.0114,
+      "step": 4878
+    },
+    {
+      "epoch": 0.9691131194756182,
+      "grad_norm": 0.677575640563524,
+      "learning_rate": 2.485120854621992e-08,
+      "loss": 0.0112,
+      "step": 4879
+    },
+    {
+      "epoch": 0.9693117489323667,
+      "grad_norm": 0.5454976845516769,
+      "learning_rate": 2.4531843806294696e-08,
+      "loss": 0.0071,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9695103783891151,
+      "grad_norm": 0.5457073272241642,
+      "learning_rate": 2.4214539402751534e-08,
+      "loss": 0.0078,
+      "step": 4881
+    },
+    {
+      "epoch": 0.9697090078458636,
+      "grad_norm": 0.4264160073509605,
+      "learning_rate": 2.3899295466983663e-08,
+      "loss": 0.0122,
+      "step": 4882
+    },
+    {
+      "epoch": 0.969907637302612,
+      "grad_norm": 0.4258970818620188,
+      "learning_rate": 2.3586112129534988e-08,
+      "loss": 0.0112,
+      "step": 4883
+    },
+    {
+      "epoch": 0.9701062667593604,
+      "grad_norm": 0.6087695074036127,
+      "learning_rate": 2.3274989520093994e-08,
+      "loss": 0.0109,
+      "step": 4884
+    },
+    {
+      "epoch": 0.9703048962161088,
+      "grad_norm": 0.857276501396098,
+      "learning_rate": 2.29659277674954e-08,
+      "loss": 0.0159,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9705035256728572,
+      "grad_norm": 0.30379709830289564,
+      "learning_rate": 2.2658926999722386e-08,
+      "loss": 0.005,
+      "step": 4886
+    },
+    {
+      "epoch": 0.9707021551296057,
+      "grad_norm": 0.44475948970448625,
+      "learning_rate": 2.2353987343902704e-08,
+      "loss": 0.0083,
+      "step": 4887
+    },
+    {
+      "epoch": 0.9709007845863542,
+      "grad_norm": 0.5243677778649878,
+      "learning_rate": 2.2051108926313125e-08,
+      "loss": 0.0063,
+      "step": 4888
+    },
+    {
+      "epoch": 0.9710994140431026,
+      "grad_norm": 0.6869080614393989,
+      "learning_rate": 2.175029187237332e-08,
+      "loss": 0.0101,
+      "step": 4889
+    },
+    {
+      "epoch": 0.971298043499851,
+      "grad_norm": 0.41495749218023437,
+      "learning_rate": 2.1451536306653088e-08,
+      "loss": 0.0096,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9714966729565995,
+      "grad_norm": 0.44501375126763126,
+      "learning_rate": 2.1154842352865134e-08,
+      "loss": 0.0076,
+      "step": 4891
+    },
+    {
+      "epoch": 0.9716953024133479,
+      "grad_norm": 0.48307549732523714,
+      "learning_rate": 2.0860210133871738e-08,
+      "loss": 0.0098,
+      "step": 4892
+    },
+    {
+      "epoch": 0.9718939318700963,
+      "grad_norm": 0.3302165574725224,
+      "learning_rate": 2.0567639771679192e-08,
+      "loss": 0.006,
+      "step": 4893
+    },
+    {
+      "epoch": 0.9720925613268447,
+      "grad_norm": 0.6477082027663088,
+      "learning_rate": 2.027713138744003e-08,
+      "loss": 0.0128,
+      "step": 4894
+    },
+    {
+      "epoch": 0.9722911907835932,
+      "grad_norm": 0.6335639722740967,
+      "learning_rate": 1.998868510145413e-08,
+      "loss": 0.0092,
+      "step": 4895
+    },
+    {
+      "epoch": 0.9724898202403417,
+      "grad_norm": 0.7004054321677288,
+      "learning_rate": 1.9702301033166505e-08,
+      "loss": 0.0134,
+      "step": 4896
+    },
+    {
+      "epoch": 0.9726884496970901,
+      "grad_norm": 0.46136836622282223,
+      "learning_rate": 1.9417979301168956e-08,
+      "loss": 0.0084,
+      "step": 4897
+    },
+    {
+      "epoch": 0.9728870791538385,
+      "grad_norm": 0.4510335143166035,
+      "learning_rate": 1.9135720023197857e-08,
+      "loss": 0.0092,
+      "step": 4898
+    },
+    {
+      "epoch": 0.973085708610587,
+      "grad_norm": 0.2875052305885156,
+      "learning_rate": 1.8855523316137492e-08,
+      "loss": 0.0056,
+      "step": 4899
+    },
+    {
+      "epoch": 0.9732843380673354,
+      "grad_norm": 0.4981975875701382,
+      "learning_rate": 1.8577389296016713e-08,
+      "loss": 0.0111,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9734829675240838,
+      "grad_norm": 0.6442750654417164,
+      "learning_rate": 1.830131807801061e-08,
+      "loss": 0.0131,
+      "step": 4901
+    },
+    {
+      "epoch": 0.9736815969808322,
+      "grad_norm": 0.4873509567237911,
+      "learning_rate": 1.802730977643996e-08,
+      "loss": 0.0079,
+      "step": 4902
+    },
+    {
+      "epoch": 0.9738802264375807,
+      "grad_norm": 0.35940737168787174,
+      "learning_rate": 1.7755364504771222e-08,
+      "loss": 0.009,
+      "step": 4903
+    },
+    {
+      "epoch": 0.9740788558943291,
+      "grad_norm": 0.6049711416734347,
+      "learning_rate": 1.7485482375616534e-08,
+      "loss": 0.0115,
+      "step": 4904
+    },
+    {
+      "epoch": 0.9742774853510776,
+      "grad_norm": 0.540096719001103,
+      "learning_rate": 1.721766350073373e-08,
+      "loss": 0.0085,
+      "step": 4905
+    },
+    {
+      "epoch": 0.974476114807826,
+      "grad_norm": 0.5403139332223295,
+      "learning_rate": 1.6951907991026863e-08,
+      "loss": 0.0104,
+      "step": 4906
+    },
+    {
+      "epoch": 0.9746747442645745,
+      "grad_norm": 0.3412169774660418,
+      "learning_rate": 1.6688215956545128e-08,
+      "loss": 0.0081,
+      "step": 4907
+    },
+    {
+      "epoch": 0.9748733737213229,
+      "grad_norm": 0.30705865819778627,
+      "learning_rate": 1.6426587506482295e-08,
+      "loss": 0.0102,
+      "step": 4908
+    },
+    {
+      "epoch": 0.9750720031780713,
+      "grad_norm": 0.46701451128719806,
+      "learning_rate": 1.616702274917892e-08,
+      "loss": 0.0101,
+      "step": 4909
+    },
+    {
+      "epoch": 0.9752706326348197,
+      "grad_norm": 0.39809981599821737,
+      "learning_rate": 1.590952179212013e-08,
+      "loss": 0.0109,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9754692620915681,
+      "grad_norm": 0.2555686289992959,
+      "learning_rate": 1.565408474193786e-08,
+      "loss": 0.0054,
+      "step": 4911
+    },
+    {
+      "epoch": 0.9756678915483166,
+      "grad_norm": 0.24199181404380177,
+      "learning_rate": 1.540071170440749e-08,
+      "loss": 0.0096,
+      "step": 4912
+    },
+    {
+      "epoch": 0.9758665210050651,
+      "grad_norm": 0.5796493821986631,
+      "learning_rate": 1.514940278445065e-08,
+      "loss": 0.0074,
+      "step": 4913
+    },
+    {
+      "epoch": 0.9760651504618135,
+      "grad_norm": 0.5748057498768764,
+      "learning_rate": 1.4900158086134097e-08,
+      "loss": 0.0154,
+      "step": 4914
+    },
+    {
+      "epoch": 0.976263779918562,
+      "grad_norm": 0.33859494804875856,
+      "learning_rate": 1.4652977712669714e-08,
+      "loss": 0.0057,
+      "step": 4915
+    },
+    {
+      "epoch": 0.9764624093753104,
+      "grad_norm": 0.4017728258831618,
+      "learning_rate": 1.4407861766415066e-08,
+      "loss": 0.0132,
+      "step": 4916
+    },
+    {
+      "epoch": 0.9766610388320588,
+      "grad_norm": 0.6787171557695976,
+      "learning_rate": 1.4164810348871739e-08,
+      "loss": 0.0118,
+      "step": 4917
+    },
+    {
+      "epoch": 0.9768596682888072,
+      "grad_norm": 0.6909520206503219,
+      "learning_rate": 1.392382356068811e-08,
+      "loss": 0.0164,
+      "step": 4918
+    },
+    {
+      "epoch": 0.9770582977455556,
+      "grad_norm": 0.5243266524129274,
+      "learning_rate": 1.3684901501655468e-08,
+      "loss": 0.0102,
+      "step": 4919
+    },
+    {
+      "epoch": 0.9772569272023041,
+      "grad_norm": 0.5013715407112671,
+      "learning_rate": 1.344804427071189e-08,
+      "loss": 0.0145,
+      "step": 4920
+    },
+    {
+      "epoch": 0.9774555566590525,
+      "grad_norm": 0.6104887396616863,
+      "learning_rate": 1.3213251965939478e-08,
+      "loss": 0.0094,
+      "step": 4921
+    },
+    {
+      "epoch": 0.977654186115801,
+      "grad_norm": 12.92537382124017,
+      "learning_rate": 1.2980524684565455e-08,
+      "loss": 0.0142,
+      "step": 4922
+    },
+    {
+      "epoch": 0.9778528155725494,
+      "grad_norm": 0.4169042250958797,
+      "learning_rate": 1.274986252296273e-08,
+      "loss": 0.0092,
+      "step": 4923
+    },
+    {
+      "epoch": 0.9780514450292979,
+      "grad_norm": 0.8118130101634882,
+      "learning_rate": 1.2521265576646569e-08,
+      "loss": 0.015,
+      "step": 4924
+    },
+    {
+      "epoch": 0.9782500744860463,
+      "grad_norm": 0.6249056291594971,
+      "learning_rate": 1.2294733940280135e-08,
+      "loss": 0.011,
+      "step": 4925
+    },
+    {
+      "epoch": 0.9784487039427947,
+      "grad_norm": 0.39444979264014,
+      "learning_rate": 1.2070267707670058e-08,
+      "loss": 0.0086,
+      "step": 4926
+    },
+    {
+      "epoch": 0.9786473333995431,
+      "grad_norm": 1.2105651014944392,
+      "learning_rate": 1.184786697176643e-08,
+      "loss": 0.0201,
+      "step": 4927
+    },
+    {
+      "epoch": 0.9788459628562916,
+      "grad_norm": 0.6828183839448554,
+      "learning_rate": 1.1627531824666138e-08,
+      "loss": 0.0137,
+      "step": 4928
+    },
+    {
+      "epoch": 0.97904459231304,
+      "grad_norm": 0.7000845033148919,
+      "learning_rate": 1.1409262357609529e-08,
+      "loss": 0.012,
+      "step": 4929
+    },
+    {
+      "epoch": 0.9792432217697885,
+      "grad_norm": 0.44350465099943603,
+      "learning_rate": 1.1193058660980971e-08,
+      "loss": 0.0115,
+      "step": 4930
+    },
+    {
+      "epoch": 0.9794418512265369,
+      "grad_norm": 0.7669306721956611,
+      "learning_rate": 1.0978920824311622e-08,
+      "loss": 0.0188,
+      "step": 4931
+    },
+    {
+      "epoch": 0.9796404806832854,
+      "grad_norm": 0.4653719983059989,
+      "learning_rate": 1.0766848936274998e-08,
+      "loss": 0.0079,
+      "step": 4932
+    },
+    {
+      "epoch": 0.9798391101400338,
+      "grad_norm": 0.3330343266105515,
+      "learning_rate": 1.0556843084689738e-08,
+      "loss": 0.0084,
+      "step": 4933
+    },
+    {
+      "epoch": 0.9800377395967822,
+      "grad_norm": 0.4031244671758501,
+      "learning_rate": 1.0348903356519057e-08,
+      "loss": 0.0088,
+      "step": 4934
+    },
+    {
+      "epoch": 0.9802363690535306,
+      "grad_norm": 0.47173933428617837,
+      "learning_rate": 1.0143029837870744e-08,
+      "loss": 0.0117,
+      "step": 4935
+    },
+    {
+      "epoch": 0.980434998510279,
+      "grad_norm": 0.5623629301817437,
+      "learning_rate": 9.939222613997157e-09,
+      "loss": 0.0088,
+      "step": 4936
+    },
+    {
+      "epoch": 0.9806336279670275,
+      "grad_norm": 0.6962143816147889,
+      "learning_rate": 9.737481769293566e-09,
+      "loss": 0.0147,
+      "step": 4937
+    },
+    {
+      "epoch": 0.980832257423776,
+      "grad_norm": 0.456040337121184,
+      "learning_rate": 9.537807387302034e-09,
+      "loss": 0.0093,
+      "step": 4938
+    },
+    {
+      "epoch": 0.9810308868805244,
+      "grad_norm": 0.4400816228735132,
+      "learning_rate": 9.340199550706974e-09,
+      "loss": 0.0062,
+      "step": 4939
+    },
+    {
+      "epoch": 0.9812295163372728,
+      "grad_norm": 0.7076270413554725,
+      "learning_rate": 9.144658341337375e-09,
+      "loss": 0.014,
+      "step": 4940
+    },
+    {
+      "epoch": 0.9814281457940213,
+      "grad_norm": 0.8596549261688607,
+      "learning_rate": 8.9511838401668e-09,
+      "loss": 0.0221,
+      "step": 4941
+    },
+    {
+      "epoch": 0.9816267752507697,
+      "grad_norm": 0.3642453441116622,
+      "learning_rate": 8.75977612731227e-09,
+      "loss": 0.0109,
+      "step": 4942
+    },
+    {
+      "epoch": 0.9818254047075181,
+      "grad_norm": 0.38678000727812256,
+      "learning_rate": 8.570435282037048e-09,
+      "loss": 0.0095,
+      "step": 4943
+    },
+    {
+      "epoch": 0.9820240341642665,
+      "grad_norm": 0.5698335800750048,
+      "learning_rate": 8.383161382745087e-09,
+      "loss": 0.0101,
+      "step": 4944
+    },
+    {
+      "epoch": 0.982222663621015,
+      "grad_norm": 0.46124466080444443,
+      "learning_rate": 8.197954506988237e-09,
+      "loss": 0.0057,
+      "step": 4945
+    },
+    {
+      "epoch": 0.9824212930777634,
+      "grad_norm": 0.3552672963335277,
+      "learning_rate": 8.014814731458487e-09,
+      "loss": 0.0059,
+      "step": 4946
+    },
+    {
+      "epoch": 0.9826199225345119,
+      "grad_norm": 0.42903428026043167,
+      "learning_rate": 7.833742131995725e-09,
+      "loss": 0.0077,
+      "step": 4947
+    },
+    {
+      "epoch": 0.9828185519912603,
+      "grad_norm": 0.5963106438933184,
+      "learning_rate": 7.65473678358053e-09,
+      "loss": 0.0132,
+      "step": 4948
+    },
+    {
+      "epoch": 0.9830171814480088,
+      "grad_norm": 0.6044799566434823,
+      "learning_rate": 7.477798760339717e-09,
+      "loss": 0.0083,
+      "step": 4949
+    },
+    {
+      "epoch": 0.9832158109047572,
+      "grad_norm": 0.25409019391898363,
+      "learning_rate": 7.302928135542453e-09,
+      "loss": 0.0096,
+      "step": 4950
+    },
+    {
+      "epoch": 0.9834144403615056,
+      "grad_norm": 0.4559853535199204,
+      "learning_rate": 7.130124981603037e-09,
+      "loss": 0.0154,
+      "step": 4951
+    },
+    {
+      "epoch": 0.983613069818254,
+      "grad_norm": 0.5446426719314116,
+      "learning_rate": 6.959389370079228e-09,
+      "loss": 0.0084,
+      "step": 4952
+    },
+    {
+      "epoch": 0.9838116992750024,
+      "grad_norm": 0.4002074352633331,
+      "learning_rate": 6.7907213716716936e-09,
+      "loss": 0.0134,
+      "step": 4953
+    },
+    {
+      "epoch": 0.9840103287317509,
+      "grad_norm": 0.5482538576918561,
+      "learning_rate": 6.624121056225674e-09,
+      "loss": 0.0101,
+      "step": 4954
+    },
+    {
+      "epoch": 0.9842089581884994,
+      "grad_norm": 0.2548625973854045,
+      "learning_rate": 6.459588492731539e-09,
+      "loss": 0.0071,
+      "step": 4955
+    },
+    {
+      "epoch": 0.9844075876452478,
+      "grad_norm": 0.7503639525229563,
+      "learning_rate": 6.297123749320344e-09,
+      "loss": 0.0075,
+      "step": 4956
+    },
+    {
+      "epoch": 0.9846062171019963,
+      "grad_norm": 0.4290624194383481,
+      "learning_rate": 6.13672689326994e-09,
+      "loss": 0.0142,
+      "step": 4957
+    },
+    {
+      "epoch": 0.9848048465587447,
+      "grad_norm": 0.47732768665483954,
+      "learning_rate": 5.978397990999973e-09,
+      "loss": 0.0084,
+      "step": 4958
+    },
+    {
+      "epoch": 0.9850034760154931,
+      "grad_norm": 0.6430141219587137,
+      "learning_rate": 5.822137108074111e-09,
+      "loss": 0.0122,
+      "step": 4959
+    },
+    {
+      "epoch": 0.9852021054722415,
+      "grad_norm": 0.4601266343624083,
+      "learning_rate": 5.6679443092000354e-09,
+      "loss": 0.0111,
+      "step": 4960
+    },
+    {
+      "epoch": 0.9854007349289899,
+      "grad_norm": 0.8545753873794333,
+      "learning_rate": 5.515819658228339e-09,
+      "loss": 0.0072,
+      "step": 4961
+    },
+    {
+      "epoch": 0.9855993643857384,
+      "grad_norm": 0.5894818362611455,
+      "learning_rate": 5.3657632181547405e-09,
+      "loss": 0.0107,
+      "step": 4962
+    },
+    {
+      "epoch": 0.9857979938424868,
+      "grad_norm": 0.6638553478242332,
+      "learning_rate": 5.217775051116203e-09,
+      "loss": 0.0135,
+      "step": 4963
+    },
+    {
+      "epoch": 0.9859966232992353,
+      "grad_norm": 0.28446784639161005,
+      "learning_rate": 5.071855218395927e-09,
+      "loss": 0.0057,
+      "step": 4964
+    },
+    {
+      "epoch": 0.9861952527559837,
+      "grad_norm": 0.5910524592127953,
+      "learning_rate": 4.9280037804178e-09,
+      "loss": 0.0144,
+      "step": 4965
+    },
+    {
+      "epoch": 0.9863938822127322,
+      "grad_norm": 0.6211297902663819,
+      "learning_rate": 4.78622079675084e-09,
+      "loss": 0.0109,
+      "step": 4966
+    },
+    {
+      "epoch": 0.9865925116694806,
+      "grad_norm": 0.5622570313549148,
+      "learning_rate": 4.64650632610808e-09,
+      "loss": 0.0115,
+      "step": 4967
+    },
+    {
+      "epoch": 0.986791141126229,
+      "grad_norm": 0.7851311388372668,
+      "learning_rate": 4.508860426344353e-09,
+      "loss": 0.0108,
+      "step": 4968
+    },
+    {
+      "epoch": 0.9869897705829774,
+      "grad_norm": 0.9651707733360702,
+      "learning_rate": 4.3732831544590625e-09,
+      "loss": 0.0158,
+      "step": 4969
+    },
+    {
+      "epoch": 0.9871884000397259,
+      "grad_norm": 1.3074665840528905,
+      "learning_rate": 4.239774566594523e-09,
+      "loss": 0.009,
+      "step": 4970
+    },
+    {
+      "epoch": 0.9873870294964743,
+      "grad_norm": 0.6168920501208438,
+      "learning_rate": 4.1083347180359555e-09,
+      "loss": 0.0158,
+      "step": 4971
+    },
+    {
+      "epoch": 0.9875856589532228,
+      "grad_norm": 0.5709471615250453,
+      "learning_rate": 3.9789636632131536e-09,
+      "loss": 0.0088,
+      "step": 4972
+    },
+    {
+      "epoch": 0.9877842884099712,
+      "grad_norm": 0.4726577740179389,
+      "learning_rate": 3.851661455698819e-09,
+      "loss": 0.0092,
+      "step": 4973
+    },
+    {
+      "epoch": 0.9879829178667197,
+      "grad_norm": 0.34416204276241585,
+      "learning_rate": 3.726428148208006e-09,
+      "loss": 0.005,
+      "step": 4974
+    },
+    {
+      "epoch": 0.9881815473234681,
+      "grad_norm": 0.5945886172045202,
+      "learning_rate": 3.6032637925997873e-09,
+      "loss": 0.0175,
+      "step": 4975
+    },
+    {
+      "epoch": 0.9883801767802165,
+      "grad_norm": 0.6385884578112276,
+      "learning_rate": 3.4821684398766987e-09,
+      "loss": 0.0105,
+      "step": 4976
+    },
+    {
+      "epoch": 0.9885788062369649,
+      "grad_norm": 0.4104801504398456,
+      "learning_rate": 3.3631421401836284e-09,
+      "loss": 0.0114,
+      "step": 4977
+    },
+    {
+      "epoch": 0.9887774356937133,
+      "grad_norm": 0.31068946253955954,
+      "learning_rate": 3.2461849428094827e-09,
+      "loss": 0.006,
+      "step": 4978
+    },
+    {
+      "epoch": 0.9889760651504618,
+      "grad_norm": 0.6688702368786089,
+      "learning_rate": 3.131296896187186e-09,
+      "loss": 0.0101,
+      "step": 4979
+    },
+    {
+      "epoch": 0.9891746946072103,
+      "grad_norm": 0.4449185151964188,
+      "learning_rate": 3.0184780478897947e-09,
+      "loss": 0.0123,
+      "step": 4980
+    },
+    {
+      "epoch": 0.9893733240639587,
+      "grad_norm": 0.5828177516464593,
+      "learning_rate": 2.907728444637159e-09,
+      "loss": 0.0127,
+      "step": 4981
+    },
+    {
+      "epoch": 0.9895719535207071,
+      "grad_norm": 0.4850738539372369,
+      "learning_rate": 2.7990481322898166e-09,
+      "loss": 0.0141,
+      "step": 4982
+    },
+    {
+      "epoch": 0.9897705829774556,
+      "grad_norm": 0.46338314092768906,
+      "learning_rate": 2.6924371558523233e-09,
+      "loss": 0.0139,
+      "step": 4983
+    },
+    {
+      "epoch": 0.989969212434204,
+      "grad_norm": 0.4030592860660505,
+      "learning_rate": 2.5878955594726974e-09,
+      "loss": 0.0117,
+      "step": 4984
+    },
+    {
+      "epoch": 0.9901678418909524,
+      "grad_norm": 1.0961667604547174,
+      "learning_rate": 2.4854233864402e-09,
+      "loss": 0.0202,
+      "step": 4985
+    },
+    {
+      "epoch": 0.9903664713477008,
+      "grad_norm": 0.4721411563769088,
+      "learning_rate": 2.3850206791897756e-09,
+      "loss": 0.0113,
+      "step": 4986
+    },
+    {
+      "epoch": 0.9905651008044493,
+      "grad_norm": 0.4833604924531587,
+      "learning_rate": 2.286687479297056e-09,
+      "loss": 0.0095,
+      "step": 4987
+    },
+    {
+      "epoch": 0.9907637302611977,
+      "grad_norm": 0.5752978520181773,
+      "learning_rate": 2.1904238274828016e-09,
+      "loss": 0.0097,
+      "step": 4988
+    },
+    {
+      "epoch": 0.9909623597179462,
+      "grad_norm": 0.39596085840021944,
+      "learning_rate": 2.0962297636084593e-09,
+      "loss": 0.0083,
+      "step": 4989
+    },
+    {
+      "epoch": 0.9911609891746946,
+      "grad_norm": 0.6351520216922857,
+      "learning_rate": 2.0041053266806054e-09,
+      "loss": 0.0117,
+      "step": 4990
+    },
+    {
+      "epoch": 0.9913596186314431,
+      "grad_norm": 0.537475588519044,
+      "learning_rate": 1.9140505548476128e-09,
+      "loss": 0.0079,
+      "step": 4991
+    },
+    {
+      "epoch": 0.9915582480881915,
+      "grad_norm": 0.4082058783031262,
+      "learning_rate": 1.8260654854013182e-09,
+      "loss": 0.0097,
+      "step": 4992
+    },
+    {
+      "epoch": 0.9917568775449399,
+      "grad_norm": 1.0157688286043665,
+      "learning_rate": 1.7401501547759104e-09,
+      "loss": 0.021,
+      "step": 4993
+    },
+    {
+      "epoch": 0.9919555070016883,
+      "grad_norm": 0.3634022299259286,
+      "learning_rate": 1.6563045985490412e-09,
+      "loss": 0.0107,
+      "step": 4994
+    },
+    {
+      "epoch": 0.9921541364584368,
+      "grad_norm": 0.6952556762857542,
+      "learning_rate": 1.5745288514407153e-09,
+      "loss": 0.0147,
+      "step": 4995
+    },
+    {
+      "epoch": 0.9923527659151852,
+      "grad_norm": 0.5786296634418221,
+      "learning_rate": 1.4948229473144005e-09,
+      "loss": 0.0117,
+      "step": 4996
+    },
+    {
+      "epoch": 0.9925513953719337,
+      "grad_norm": 0.6356775912475472,
+      "learning_rate": 1.417186919176472e-09,
+      "loss": 0.0151,
+      "step": 4997
+    },
+    {
+      "epoch": 0.9927500248286821,
+      "grad_norm": 0.5692096365130425,
+      "learning_rate": 1.341620799175658e-09,
+      "loss": 0.0076,
+      "step": 4998
+    },
+    {
+      "epoch": 0.9929486542854306,
+      "grad_norm": 0.8015252962881247,
+      "learning_rate": 1.2681246186035945e-09,
+      "loss": 0.0166,
+      "step": 4999
+    },
+    {
+      "epoch": 0.993147283742179,
+      "grad_norm": 0.5276135206707997,
+      "learning_rate": 1.1966984078959354e-09,
+      "loss": 0.0087,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9933459131989274,
+      "grad_norm": 0.4908418867095688,
+      "learning_rate": 1.1273421966290221e-09,
+      "loss": 0.0094,
+      "step": 5001
+    },
+    {
+      "epoch": 0.9935445426556758,
+      "grad_norm": 0.9923607226693753,
+      "learning_rate": 1.0600560135237691e-09,
+      "loss": 0.0145,
+      "step": 5002
+    },
+    {
+      "epoch": 0.9937431721124242,
+      "grad_norm": 0.2731344484135086,
+      "learning_rate": 9.948398864434439e-10,
+      "loss": 0.0068,
+      "step": 5003
+    },
+    {
+      "epoch": 0.9939418015691727,
+      "grad_norm": 0.7287487195663518,
+      "learning_rate": 9.316938423936662e-10,
+      "loss": 0.0092,
+      "step": 5004
+    },
+    {
+      "epoch": 0.9941404310259211,
+      "grad_norm": 0.505240386716718,
+      "learning_rate": 8.706179075229637e-10,
+      "loss": 0.0117,
+      "step": 5005
+    },
+    {
+      "epoch": 0.9943390604826696,
+      "grad_norm": 0.5101453671929035,
+      "learning_rate": 8.116121071238825e-10,
+      "loss": 0.0136,
+      "step": 5006
+    },
+    {
+      "epoch": 0.994537689939418,
+      "grad_norm": 0.4807315646835647,
+      "learning_rate": 7.546764656291005e-10,
+      "loss": 0.0086,
+      "step": 5007
+    },
+    {
+      "epoch": 0.9947363193961665,
+      "grad_norm": 0.4595127871163573,
+      "learning_rate": 6.998110066169794e-10,
+      "loss": 0.0083,
+      "step": 5008
+    },
+    {
+      "epoch": 0.9949349488529149,
+      "grad_norm": 0.5817190396655636,
+      "learning_rate": 6.470157528065679e-10,
+      "loss": 0.0101,
+      "step": 5009
+    },
+    {
+      "epoch": 0.9951335783096633,
+      "grad_norm": 0.3676483776227927,
+      "learning_rate": 5.962907260603779e-10,
+      "loss": 0.0088,
+      "step": 5010
+    },
+    {
+      "epoch": 0.9953322077664117,
+      "grad_norm": 0.5527914548170548,
+      "learning_rate": 5.476359473838289e-10,
+      "loss": 0.0089,
+      "step": 5011
+    },
+    {
+      "epoch": 0.9955308372231602,
+      "grad_norm": 0.47874706302938747,
+      "learning_rate": 5.010514369246933e-10,
+      "loss": 0.0131,
+      "step": 5012
+    },
+    {
+      "epoch": 0.9957294666799086,
+      "grad_norm": 0.4328300045576964,
+      "learning_rate": 4.565372139730961e-10,
+      "loss": 0.0162,
+      "step": 5013
+    },
+    {
+      "epoch": 0.9959280961366571,
+      "grad_norm": 0.45665840240784356,
+      "learning_rate": 4.140932969631806e-10,
+      "loss": 0.0085,
+      "step": 5014
+    },
+    {
+      "epoch": 0.9961267255934055,
+      "grad_norm": 0.9001422570931706,
+      "learning_rate": 3.737197034703322e-10,
+      "loss": 0.0126,
+      "step": 5015
+    },
+    {
+      "epoch": 0.996325355050154,
+      "grad_norm": 0.45807153798940137,
+      "learning_rate": 3.3541645021339944e-10,
+      "loss": 0.0105,
+      "step": 5016
+    },
+    {
+      "epoch": 0.9965239845069024,
+      "grad_norm": 0.6282268630028757,
+      "learning_rate": 2.991835530535836e-10,
+      "loss": 0.0092,
+      "step": 5017
+    },
+    {
+      "epoch": 0.9967226139636508,
+      "grad_norm": 0.42023530746133536,
+      "learning_rate": 2.650210269955489e-10,
+      "loss": 0.01,
+      "step": 5018
+    },
+    {
+      "epoch": 0.9969212434203992,
+      "grad_norm": 0.40516998628725354,
+      "learning_rate": 2.3292888618520195e-10,
+      "loss": 0.006,
+      "step": 5019
+    },
+    {
+      "epoch": 0.9971198728771477,
+      "grad_norm": 0.5106870738378689,
+      "learning_rate": 2.0290714391191235e-10,
+      "loss": 0.0135,
+      "step": 5020
+    },
+    {
+      "epoch": 0.9973185023338961,
+      "grad_norm": 0.5702272196550405,
+      "learning_rate": 1.7495581260795758e-10,
+      "loss": 0.0084,
+      "step": 5021
+    },
+    {
+      "epoch": 0.9975171317906446,
+      "grad_norm": 0.3574001167310024,
+      "learning_rate": 1.4907490384796774e-10,
+      "loss": 0.0089,
+      "step": 5022
+    },
+    {
+      "epoch": 0.997715761247393,
+      "grad_norm": 0.6482126256169631,
+      "learning_rate": 1.252644283489257e-10,
+      "loss": 0.009,
+      "step": 5023
+    },
+    {
+      "epoch": 0.9979143907041415,
+      "grad_norm": 0.40764995367026596,
+      "learning_rate": 1.0352439597072217e-10,
+      "loss": 0.0123,
+      "step": 5024
+    },
+    {
+      "epoch": 0.9981130201608899,
+      "grad_norm": 0.464525483618827,
+      "learning_rate": 8.385481571615561e-11,
+      "loss": 0.0124,
+      "step": 5025
+    },
+    {
+      "epoch": 0.9983116496176383,
+      "grad_norm": 0.49823901728048225,
+      "learning_rate": 6.625569573037727e-11,
+      "loss": 0.0176,
+      "step": 5026
+    },
+    {
+      "epoch": 0.9985102790743867,
+      "grad_norm": 0.5896583089333627,
+      "learning_rate": 5.0727043301446175e-11,
+      "loss": 0.0129,
+      "step": 5027
+    },
+    {
+      "epoch": 0.9987089085311351,
+      "grad_norm": 0.37718748502681293,
+      "learning_rate": 3.726886485866388e-11,
+      "loss": 0.0133,
+      "step": 5028
+    },
+    {
+      "epoch": 0.9989075379878836,
+      "grad_norm": 0.5830988240086593,
+      "learning_rate": 2.5881165976460178e-11,
+      "loss": 0.0161,
+      "step": 5029
+    },
+    {
+      "epoch": 0.999106167444632,
+      "grad_norm": 0.6251502587104545,
+      "learning_rate": 1.6563951368842034e-11,
+      "loss": 0.0125,
+      "step": 5030
+    },
+    {
+      "epoch": 0.9993047969013805,
+      "grad_norm": 0.5850394153814858,
+      "learning_rate": 9.317224895499799e-12,
+      "loss": 0.013,
+      "step": 5031
+    },
+    {
+      "epoch": 0.9995034263581289,
+      "grad_norm": 0.5740375396508653,
+      "learning_rate": 4.1409895568111924e-12,
+      "loss": 0.0112,
+      "step": 5032
+    },
+    {
+      "epoch": 0.9997020558148774,
+      "grad_norm": 0.2949227457860251,
+      "learning_rate": 1.0352474966168758e-12,
+      "loss": 0.0055,
+      "step": 5033
+    },
+    {
+      "epoch": 0.9999006852716258,
+      "grad_norm": 0.8152259997227421,
+      "learning_rate": 0.0,
+      "loss": 0.0059,
+      "step": 5034
+    },
+    {
+      "epoch": 0.9999006852716258,
+      "step": 5034,
+      "total_flos": 393450150948864.0,
+      "train_loss": 0.013911830744355322,
+      "train_runtime": 43586.62,
+      "train_samples_per_second": 7.393,
+      "train_steps_per_second": 0.115
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5034,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 393450150948864.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}