diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,20484 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 200,
+  "global_step": 2836,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 202.90386993967232,
+      "learning_rate": 1.1627906976744187e-07,
+      "loss": 1.8438,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 228.39819277358856,
+      "learning_rate": 2.3255813953488374e-07,
+      "loss": 1.7363,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 214.780561941506,
+      "learning_rate": 3.488372093023256e-07,
+      "loss": 1.7539,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 216.4901890000185,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 1.7754,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 231.69852346374375,
+      "learning_rate": 5.813953488372094e-07,
+      "loss": 1.7812,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 197.05129366480213,
+      "learning_rate": 6.976744186046513e-07,
+      "loss": 1.4727,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 197.1596840015841,
+      "learning_rate": 8.139534883720931e-07,
+      "loss": 1.373,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 124.2201938585486,
+      "learning_rate": 9.30232558139535e-07,
+      "loss": 1.0439,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 103.7999847150281,
+      "learning_rate": 1.0465116279069768e-06,
+      "loss": 0.8701,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.475338111839356,
+      "learning_rate": 1.1627906976744188e-06,
+      "loss": 0.7646,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 44.68636332290023,
+      "learning_rate": 1.2790697674418605e-06,
+      "loss": 0.8047,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 98.57286895249709,
+      "learning_rate": 1.3953488372093025e-06,
+      "loss": 1.4346,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 57.79959616068188,
+      "learning_rate": 1.5116279069767443e-06,
+      "loss": 1.0117,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 70.09256168636482,
+      "learning_rate": 1.6279069767441862e-06,
+      "loss": 1.166,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 61.157329357550594,
+      "learning_rate": 1.7441860465116282e-06,
+      "loss": 1.0713,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 71.47010478235612,
+      "learning_rate": 1.86046511627907e-06,
+      "loss": 1.5044,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 57.6605998963668,
+      "learning_rate": 1.976744186046512e-06,
+      "loss": 1.1948,
+      "step": 17
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 62.342349612174615,
+      "learning_rate": 2.0930232558139536e-06,
+      "loss": 1.293,
+      "step": 18
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 98.38003497453664,
+      "learning_rate": 2.2093023255813954e-06,
+      "loss": 2.123,
+      "step": 19
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 59.080195992834504,
+      "learning_rate": 2.3255813953488376e-06,
+      "loss": 1.2031,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 30.029214772660623,
+      "learning_rate": 2.4418604651162793e-06,
+      "loss": 0.7253,
+      "step": 21
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 37.0121588071573,
+      "learning_rate": 2.558139534883721e-06,
+      "loss": 0.7798,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 29.58311274643164,
+      "learning_rate": 2.674418604651163e-06,
+      "loss": 0.6982,
+      "step": 23
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.304958829474334,
+      "learning_rate": 2.790697674418605e-06,
+      "loss": 0.564,
+      "step": 24
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.297286194315642,
+      "learning_rate": 2.9069767441860468e-06,
+      "loss": 0.5098,
+      "step": 25
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 43.53145865272809,
+      "learning_rate": 3.0232558139534885e-06,
+      "loss": 0.623,
+      "step": 26
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 29.257611801383778,
+      "learning_rate": 3.1395348837209307e-06,
+      "loss": 0.5215,
+      "step": 27
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 33.17177984579731,
+      "learning_rate": 3.2558139534883724e-06,
+      "loss": 0.5112,
+      "step": 28
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.909974521793234,
+      "learning_rate": 3.372093023255814e-06,
+      "loss": 0.417,
+      "step": 29
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.64535376272361,
+      "learning_rate": 3.4883720930232564e-06,
+      "loss": 0.4351,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.979845496414617,
+      "learning_rate": 3.6046511627906977e-06,
+      "loss": 0.4712,
+      "step": 31
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.131386941035599,
+      "learning_rate": 3.72093023255814e-06,
+      "loss": 0.3901,
+      "step": 32
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.9258120439031,
+      "learning_rate": 3.837209302325582e-06,
+      "loss": 0.5146,
+      "step": 33
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.818651310911143,
+      "learning_rate": 3.953488372093024e-06,
+      "loss": 0.3828,
+      "step": 34
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.348280931060449,
+      "learning_rate": 4.0697674418604655e-06,
+      "loss": 0.3921,
+      "step": 35
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.276575722764843,
+      "learning_rate": 4.186046511627907e-06,
+      "loss": 0.3508,
+      "step": 36
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.874510152258896,
+      "learning_rate": 4.302325581395349e-06,
+      "loss": 0.3513,
+      "step": 37
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.694474088512809,
+      "learning_rate": 4.418604651162791e-06,
+      "loss": 0.3115,
+      "step": 38
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.188014836504826,
+      "learning_rate": 4.5348837209302326e-06,
+      "loss": 0.3945,
+      "step": 39
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.398178396756387,
+      "learning_rate": 4.651162790697675e-06,
+      "loss": 0.4326,
+      "step": 40
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.126489213952297,
+      "learning_rate": 4.767441860465117e-06,
+      "loss": 0.3462,
+      "step": 41
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.672306859157442,
+      "learning_rate": 4.883720930232559e-06,
+      "loss": 0.3337,
+      "step": 42
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.641764743791803,
+      "learning_rate": 5e-06,
+      "loss": 0.364,
+      "step": 43
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.1660458873902435,
+      "learning_rate": 5.116279069767442e-06,
+      "loss": 0.416,
+      "step": 44
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.068038045757105,
+      "learning_rate": 5.232558139534885e-06,
+      "loss": 0.3359,
+      "step": 45
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.885466229956409,
+      "learning_rate": 5.348837209302326e-06,
+      "loss": 0.3093,
+      "step": 46
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.402483663272646,
+      "learning_rate": 5.465116279069767e-06,
+      "loss": 0.4321,
+      "step": 47
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.254378914329732,
+      "learning_rate": 5.58139534883721e-06,
+      "loss": 0.3677,
+      "step": 48
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.102510647971088,
+      "learning_rate": 5.697674418604652e-06,
+      "loss": 0.3315,
+      "step": 49
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.103089200793725,
+      "learning_rate": 5.8139534883720935e-06,
+      "loss": 0.3579,
+      "step": 50
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.896709830989335,
+      "learning_rate": 5.930232558139536e-06,
+      "loss": 0.3015,
+      "step": 51
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.038774423193659,
+      "learning_rate": 6.046511627906977e-06,
+      "loss": 0.4043,
+      "step": 52
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.767550745321053,
+      "learning_rate": 6.162790697674419e-06,
+      "loss": 0.4272,
+      "step": 53
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.104250876307582,
+      "learning_rate": 6.279069767441861e-06,
+      "loss": 0.4619,
+      "step": 54
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.516390000248048,
+      "learning_rate": 6.395348837209303e-06,
+      "loss": 0.3489,
+      "step": 55
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.041894730595576,
+      "learning_rate": 6.511627906976745e-06,
+      "loss": 0.4575,
+      "step": 56
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.34509202786394,
+      "learning_rate": 6.627906976744186e-06,
+      "loss": 0.4976,
+      "step": 57
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.251549794051696,
+      "learning_rate": 6.744186046511628e-06,
+      "loss": 0.3945,
+      "step": 58
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.25510184244114,
+      "learning_rate": 6.86046511627907e-06,
+      "loss": 0.4683,
+      "step": 59
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 19.283905016519853,
+      "learning_rate": 6.976744186046513e-06,
+      "loss": 0.355,
+      "step": 60
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.112124591879601,
+      "learning_rate": 7.0930232558139545e-06,
+      "loss": 0.4082,
+      "step": 61
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.508937939078336,
+      "learning_rate": 7.209302325581395e-06,
+      "loss": 0.3147,
+      "step": 62
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.642074863247307,
+      "learning_rate": 7.325581395348837e-06,
+      "loss": 0.376,
+      "step": 63
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.42838089389136,
+      "learning_rate": 7.44186046511628e-06,
+      "loss": 0.3364,
+      "step": 64
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.6088894091488335,
+      "learning_rate": 7.5581395348837215e-06,
+      "loss": 0.2657,
+      "step": 65
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.219539117606181,
+      "learning_rate": 7.674418604651164e-06,
+      "loss": 0.2565,
+      "step": 66
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 22.748349938573973,
+      "learning_rate": 7.790697674418605e-06,
+      "loss": 0.4777,
+      "step": 67
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.743555684373569,
+      "learning_rate": 7.906976744186048e-06,
+      "loss": 0.2827,
+      "step": 68
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.60049706316629,
+      "learning_rate": 8.023255813953488e-06,
+      "loss": 0.3711,
+      "step": 69
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.538346266590203,
+      "learning_rate": 8.139534883720931e-06,
+      "loss": 0.3474,
+      "step": 70
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.71151587878829,
+      "learning_rate": 8.255813953488374e-06,
+      "loss": 0.2908,
+      "step": 71
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.402535865702138,
+      "learning_rate": 8.372093023255815e-06,
+      "loss": 0.376,
+      "step": 72
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.986376366002659,
+      "learning_rate": 8.488372093023256e-06,
+      "loss": 0.3408,
+      "step": 73
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 20.48661268694837,
+      "learning_rate": 8.604651162790698e-06,
+      "loss": 0.4917,
+      "step": 74
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.211950604192614,
+      "learning_rate": 8.72093023255814e-06,
+      "loss": 0.2549,
+      "step": 75
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.323294237384195,
+      "learning_rate": 8.837209302325582e-06,
+      "loss": 0.3533,
+      "step": 76
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.72312055968665,
+      "learning_rate": 8.953488372093024e-06,
+      "loss": 0.3499,
+      "step": 77
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.920434602536187,
+      "learning_rate": 9.069767441860465e-06,
+      "loss": 0.4331,
+      "step": 78
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.133336754291587,
+      "learning_rate": 9.186046511627908e-06,
+      "loss": 0.2417,
+      "step": 79
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.04857063542844,
+      "learning_rate": 9.30232558139535e-06,
+      "loss": 0.3987,
+      "step": 80
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.498206047945851,
+      "learning_rate": 9.418604651162791e-06,
+      "loss": 0.3887,
+      "step": 81
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.43711253894265,
+      "learning_rate": 9.534883720930234e-06,
+      "loss": 0.4834,
+      "step": 82
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.715908189886997,
+      "learning_rate": 9.651162790697676e-06,
+      "loss": 0.4341,
+      "step": 83
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 31.158933041547197,
+      "learning_rate": 9.767441860465117e-06,
+      "loss": 0.3857,
+      "step": 84
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.603957809890236,
+      "learning_rate": 9.883720930232558e-06,
+      "loss": 0.3564,
+      "step": 85
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.294802609744977,
+      "learning_rate": 1e-05,
+      "loss": 0.3169,
+      "step": 86
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.834791745311425,
+      "learning_rate": 9.999996737321215e-06,
+      "loss": 0.3127,
+      "step": 87
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.5613775566577885,
+      "learning_rate": 9.999986949289115e-06,
+      "loss": 0.2878,
+      "step": 88
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.250929756210054,
+      "learning_rate": 9.999970635916476e-06,
+      "loss": 0.3618,
+      "step": 89
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.08442432281657,
+      "learning_rate": 9.999947797224585e-06,
+      "loss": 0.4697,
+      "step": 90
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.153753755001682,
+      "learning_rate": 9.999918433243253e-06,
+      "loss": 0.3398,
+      "step": 91
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.951376361860433,
+      "learning_rate": 9.999882544010798e-06,
+      "loss": 0.3491,
+      "step": 92
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.82012402993495,
+      "learning_rate": 9.999840129574059e-06,
+      "loss": 0.3667,
+      "step": 93
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.784955263448666,
+      "learning_rate": 9.99979118998839e-06,
+      "loss": 0.3188,
+      "step": 94
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.365867320076399,
+      "learning_rate": 9.999735725317661e-06,
+      "loss": 0.4453,
+      "step": 95
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.597515794071233,
+      "learning_rate": 9.999673735634259e-06,
+      "loss": 0.3872,
+      "step": 96
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.703300688204541,
+      "learning_rate": 9.999605221019082e-06,
+      "loss": 0.3188,
+      "step": 97
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.836799912628004,
+      "learning_rate": 9.999530181561548e-06,
+      "loss": 0.3682,
+      "step": 98
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.470745311145016,
+      "learning_rate": 9.999448617359588e-06,
+      "loss": 0.2974,
+      "step": 99
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.658427995468918,
+      "learning_rate": 9.99936052851965e-06,
+      "loss": 0.3518,
+      "step": 100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.627088136026523,
+      "learning_rate": 9.999265915156697e-06,
+      "loss": 0.2991,
+      "step": 101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.345912004866994,
+      "learning_rate": 9.999164777394204e-06,
+      "loss": 0.2405,
+      "step": 102
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.115071699429476,
+      "learning_rate": 9.999057115364166e-06,
+      "loss": 0.3242,
+      "step": 103
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.148856233512934,
+      "learning_rate": 9.998942929207085e-06,
+      "loss": 0.2302,
+      "step": 104
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.685789584085792,
+      "learning_rate": 9.998822219071987e-06,
+      "loss": 0.3203,
+      "step": 105
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.87440606933403,
+      "learning_rate": 9.998694985116406e-06,
+      "loss": 0.2732,
+      "step": 106
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.558562603253549,
+      "learning_rate": 9.998561227506388e-06,
+      "loss": 0.3599,
+      "step": 107
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.913629530029457,
+      "learning_rate": 9.9984209464165e-06,
+      "loss": 0.4126,
+      "step": 108
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.513039439858822,
+      "learning_rate": 9.99827414202982e-06,
+      "loss": 0.4067,
+      "step": 109
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.533606461587016,
+      "learning_rate": 9.998120814537932e-06,
+      "loss": 0.3279,
+      "step": 110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.789148874178633,
+      "learning_rate": 9.997960964140946e-06,
+      "loss": 0.3696,
+      "step": 111
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.4087759747852955,
+      "learning_rate": 9.997794591047476e-06,
+      "loss": 0.3354,
+      "step": 112
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.520164904897284,
+      "learning_rate": 9.99762169547465e-06,
+      "loss": 0.3472,
+      "step": 113
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.228879404608895,
+      "learning_rate": 9.997442277648108e-06,
+      "loss": 0.3506,
+      "step": 114
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.78368121969857,
+      "learning_rate": 9.997256337802006e-06,
+      "loss": 0.2859,
+      "step": 115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.979858078968924,
+      "learning_rate": 9.997063876179007e-06,
+      "loss": 0.4016,
+      "step": 116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.264068685658629,
+      "learning_rate": 9.996864893030288e-06,
+      "loss": 0.3667,
+      "step": 117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.544315637398817,
+      "learning_rate": 9.996659388615535e-06,
+      "loss": 0.4805,
+      "step": 118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.875340698729603,
+      "learning_rate": 9.996447363202947e-06,
+      "loss": 0.4111,
+      "step": 119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.533535773702745,
+      "learning_rate": 9.996228817069232e-06,
+      "loss": 0.4463,
+      "step": 120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.03323774156283,
+      "learning_rate": 9.996003750499608e-06,
+      "loss": 0.4268,
+      "step": 121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.511817295313715,
+      "learning_rate": 9.995772163787805e-06,
+      "loss": 0.3799,
+      "step": 122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.311450957665407,
+      "learning_rate": 9.995534057236057e-06,
+      "loss": 0.3467,
+      "step": 123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.487138467305234,
+      "learning_rate": 9.99528943115511e-06,
+      "loss": 0.3477,
+      "step": 124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.32037306730503,
+      "learning_rate": 9.995038285864223e-06,
+      "loss": 0.3118,
+      "step": 125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.566194177448926,
+      "learning_rate": 9.994780621691156e-06,
+      "loss": 0.3982,
+      "step": 126
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.579528821891046,
+      "learning_rate": 9.994516438972176e-06,
+      "loss": 0.4146,
+      "step": 127
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.537015286566316,
+      "learning_rate": 9.994245738052067e-06,
+      "loss": 0.3501,
+      "step": 128
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.37040147283192,
+      "learning_rate": 9.993968519284106e-06,
+      "loss": 0.4058,
+      "step": 129
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.1342384222658195,
+      "learning_rate": 9.99368478303009e-06,
+      "loss": 0.3718,
+      "step": 130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.463482414265249,
+      "learning_rate": 9.993394529660307e-06,
+      "loss": 0.3994,
+      "step": 131
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.955532663273292,
+      "learning_rate": 9.993097759553567e-06,
+      "loss": 0.4116,
+      "step": 132
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.901189275612022,
+      "learning_rate": 9.992794473097172e-06,
+      "loss": 0.4888,
+      "step": 133
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.798607971917969,
+      "learning_rate": 9.992484670686931e-06,
+      "loss": 0.3511,
+      "step": 134
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.277920259587939,
+      "learning_rate": 9.99216835272716e-06,
+      "loss": 0.3335,
+      "step": 135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.939114561364981,
+      "learning_rate": 9.991845519630679e-06,
+      "loss": 0.406,
+      "step": 136
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.408817395467018,
+      "learning_rate": 9.991516171818805e-06,
+      "loss": 0.3357,
+      "step": 137
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.57996104293568,
+      "learning_rate": 9.991180309721362e-06,
+      "loss": 0.3164,
+      "step": 138
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.202683289679439,
+      "learning_rate": 9.990837933776672e-06,
+      "loss": 0.2888,
+      "step": 139
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.629942025775903,
+      "learning_rate": 9.990489044431562e-06,
+      "loss": 0.3586,
+      "step": 140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5932529816892296,
+      "learning_rate": 9.990133642141359e-06,
+      "loss": 0.2615,
+      "step": 141
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4267135080096147,
+      "learning_rate": 9.989771727369885e-06,
+      "loss": 0.272,
+      "step": 142
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.544810985761825,
+      "learning_rate": 9.989403300589466e-06,
+      "loss": 0.3162,
+      "step": 143
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.281370620955446,
+      "learning_rate": 9.989028362280927e-06,
+      "loss": 0.281,
+      "step": 144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.5600461653530395,
+      "learning_rate": 9.988646912933586e-06,
+      "loss": 0.3584,
+      "step": 145
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.5683149127936025,
+      "learning_rate": 9.988258953045264e-06,
+      "loss": 0.3669,
+      "step": 146
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.252741698933907,
+      "learning_rate": 9.987864483122277e-06,
+      "loss": 0.3704,
+      "step": 147
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.194887530478056,
+      "learning_rate": 9.987463503679434e-06,
+      "loss": 0.3572,
+      "step": 148
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.997196544497026,
+      "learning_rate": 9.987056015240045e-06,
+      "loss": 0.4097,
+      "step": 149
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.402384927130485,
+      "learning_rate": 9.986642018335907e-06,
+      "loss": 0.3442,
+      "step": 150
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.8431969002019715,
+      "learning_rate": 9.98622151350732e-06,
+      "loss": 0.4194,
+      "step": 151
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.122840263791073,
+      "learning_rate": 9.98579450130307e-06,
+      "loss": 0.3066,
+      "step": 152
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.45755016225541,
+      "learning_rate": 9.985360982280441e-06,
+      "loss": 0.3677,
+      "step": 153
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.673939922811587,
+      "learning_rate": 9.984920957005205e-06,
+      "loss": 0.3052,
+      "step": 154
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.621777913405918,
+      "learning_rate": 9.984474426051626e-06,
+      "loss": 0.4619,
+      "step": 155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.640966980645174,
+      "learning_rate": 9.984021390002458e-06,
+      "loss": 0.3164,
+      "step": 156
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.839301744962934,
+      "learning_rate": 9.983561849448948e-06,
+      "loss": 0.3635,
+      "step": 157
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.665353230347845,
+      "learning_rate": 9.983095804990828e-06,
+      "loss": 0.4888,
+      "step": 158
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.924730676250316,
+      "learning_rate": 9.98262325723632e-06,
+      "loss": 0.3047,
+      "step": 159
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.055279988071614,
+      "learning_rate": 9.982144206802131e-06,
+      "loss": 0.3022,
+      "step": 160
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.311615501701311,
+      "learning_rate": 9.981658654313458e-06,
+      "loss": 0.3301,
+      "step": 161
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.650937395921243,
+      "learning_rate": 9.981166600403979e-06,
+      "loss": 0.2549,
+      "step": 162
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.956841091161687,
+      "learning_rate": 9.980668045715864e-06,
+      "loss": 0.3521,
+      "step": 163
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.810248909172863,
+      "learning_rate": 9.980162990899758e-06,
+      "loss": 0.3794,
+      "step": 164
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 20.022790019419208,
+      "learning_rate": 9.979651436614794e-06,
+      "loss": 0.4424,
+      "step": 165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.128512112184176,
+      "learning_rate": 9.979133383528591e-06,
+      "loss": 0.3425,
+      "step": 166
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.740296670051944,
+      "learning_rate": 9.978608832317242e-06,
+      "loss": 0.3357,
+      "step": 167
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.194690302271221,
+      "learning_rate": 9.978077783665324e-06,
+      "loss": 0.3118,
+      "step": 168
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.108082978743148,
+      "learning_rate": 9.977540238265894e-06,
+      "loss": 0.2485,
+      "step": 169
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.912776899674932,
+      "learning_rate": 9.976996196820487e-06,
+      "loss": 0.4258,
+      "step": 170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.464702725255747,
+      "learning_rate": 9.976445660039118e-06,
+      "loss": 0.345,
+      "step": 171
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.442280598388425,
+      "learning_rate": 9.975888628640273e-06,
+      "loss": 0.4453,
+      "step": 172
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.897540433653885,
+      "learning_rate": 9.97532510335092e-06,
+      "loss": 0.2983,
+      "step": 173
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.74853854756219,
+      "learning_rate": 9.974755084906503e-06,
+      "loss": 0.3584,
+      "step": 174
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.262866920229017,
+      "learning_rate": 9.97417857405093e-06,
+      "loss": 0.313,
+      "step": 175
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.617500683444464,
+      "learning_rate": 9.973595571536593e-06,
+      "loss": 0.3335,
+      "step": 176
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.94388264655118,
+      "learning_rate": 9.973006078124352e-06,
+      "loss": 0.3022,
+      "step": 177
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.806238064762597,
+      "learning_rate": 9.972410094583538e-06,
+      "loss": 0.3191,
+      "step": 178
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.69212230864316,
+      "learning_rate": 9.971807621691948e-06,
+      "loss": 0.3807,
+      "step": 179
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.009711594651918,
+      "learning_rate": 9.971198660235862e-06,
+      "loss": 0.3538,
+      "step": 180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.0933110999032225,
+      "learning_rate": 9.970583211010008e-06,
+      "loss": 0.343,
+      "step": 181
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.647852312315236,
+      "learning_rate": 9.969961274817596e-06,
+      "loss": 0.3179,
+      "step": 182
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.178827505947593,
+      "learning_rate": 9.969332852470296e-06,
+      "loss": 0.3833,
+      "step": 183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.762339182571518,
+      "learning_rate": 9.968697944788246e-06,
+      "loss": 0.2666,
+      "step": 184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.846445669213327,
+      "learning_rate": 9.968056552600043e-06,
+      "loss": 0.3433,
+      "step": 185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.72749574357049,
+      "learning_rate": 9.96740867674275e-06,
+      "loss": 0.3171,
+      "step": 186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.461494158435758,
+      "learning_rate": 9.966754318061897e-06,
+      "loss": 0.3428,
+      "step": 187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.85540674897898,
+      "learning_rate": 9.96609347741146e-06,
+      "loss": 0.3315,
+      "step": 188
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.661042480040287,
+      "learning_rate": 9.96542615565389e-06,
+      "loss": 0.3301,
+      "step": 189
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.312780660015604,
+      "learning_rate": 9.964752353660088e-06,
+      "loss": 0.2703,
+      "step": 190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.639942089377154,
+      "learning_rate": 9.964072072309412e-06,
+      "loss": 0.3865,
+      "step": 191
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.21019404742534,
+      "learning_rate": 9.96338531248968e-06,
+      "loss": 0.3167,
+      "step": 192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.096032933084754,
+      "learning_rate": 9.96269207509716e-06,
+      "loss": 0.3384,
+      "step": 193
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.15586712480747,
+      "learning_rate": 9.96199236103658e-06,
+      "loss": 0.2922,
+      "step": 194
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.043490714442856,
+      "learning_rate": 9.961286171221113e-06,
+      "loss": 0.301,
+      "step": 195
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.162555123920532,
+      "learning_rate": 9.960573506572391e-06,
+      "loss": 0.425,
+      "step": 196
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.575963914191476,
+      "learning_rate": 9.959854368020488e-06,
+      "loss": 0.3975,
+      "step": 197
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.707269520658982,
+      "learning_rate": 9.959128756503935e-06,
+      "loss": 0.291,
+      "step": 198
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.858214079166844,
+      "learning_rate": 9.958396672969706e-06,
+      "loss": 0.3721,
+      "step": 199
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.780407670857627,
+      "learning_rate": 9.957658118373222e-06,
+      "loss": 0.3965,
+      "step": 200
+    },
+    {
+      "epoch": 0.28,
+      "eval_avg_AUC": 0.7353366695144332,
+      "eval_avg_Accuracy": 0.6511521883289124,
+      "eval_avg_Accuracy-right": 0.9609364810225642,
+      "eval_avg_Accuracy-wrong": 0.1109847623379577,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6234717363160045,
+      "eval_last_AUC": 0.7578732523292513,
+      "eval_last_Accuracy": 0.7063577586206896,
+      "eval_last_Accuracy-right": 0.8229424807617061,
+      "eval_last_Accuracy-wrong": 0.5030702751876279,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6308774816172676,
+      "eval_max_AUC": 0.6025924113394245,
+      "eval_max_Accuracy": 0.6330818965517241,
+      "eval_max_Accuracy-right": 0.9913916786226685,
+      "eval_max_Accuracy-wrong": 0.008301114396179213,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5394135494496163,
+      "eval_min_AUC": 0.7575919302022417,
+      "eval_min_Accuracy": 0.7103365384615384,
+      "eval_min_Accuracy-right": 0.8019433937654885,
+      "eval_min_Accuracy-wrong": 0.5506026836479417,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6351338830995708,
+      "eval_prod_AUC": 0.750514285814293,
+      "eval_prod_Accuracy": 0.6252901193633952,
+      "eval_prod_Accuracy-right": 0.5002608582235555,
+      "eval_prod_Accuracy-wrong": 0.8433022515351376,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6219659103231394,
+      "eval_runtime": 669.486,
+      "eval_samples_per_second": 36.04,
+      "eval_steps_per_second": 1.126,
+      "eval_sum_AUC": 0.561136554889877,
+      "eval_sum_Accuracy": 0.6343252652519894,
+      "eval_sum_Accuracy-right": 0.9976522759880005,
+      "eval_sum_Accuracy-wrong": 0.0007959972708664999,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.5999569030849772,
+      "step": 200
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.335716781496041,
+      "learning_rate": 9.95691309367835e-06,
+      "loss": 0.3633,
+      "step": 201
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 16.6255160622814,
+      "learning_rate": 9.956161599857398e-06,
+      "loss": 0.4448,
+      "step": 202
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.768020408991362,
+      "learning_rate": 9.955403637891122e-06,
+      "loss": 0.4141,
+      "step": 203
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.184087162278392,
+      "learning_rate": 9.954639208768717e-06,
+      "loss": 0.3242,
+      "step": 204
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.731992970955969,
+      "learning_rate": 9.953868313487816e-06,
+      "loss": 0.395,
+      "step": 205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.952092629271191,
+      "learning_rate": 9.953090953054491e-06,
+      "loss": 0.3311,
+      "step": 206
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.4163884628076575,
+      "learning_rate": 9.952307128483257e-06,
+      "loss": 0.3311,
+      "step": 207
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.860795589168655,
+      "learning_rate": 9.951516840797058e-06,
+      "loss": 0.4414,
+      "step": 208
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.594748961130638,
+      "learning_rate": 9.950720091027277e-06,
+      "loss": 0.4131,
+      "step": 209
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.827137422936143,
+      "learning_rate": 9.94991688021373e-06,
+      "loss": 0.3247,
+      "step": 210
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.203676630208802,
+      "learning_rate": 9.949107209404664e-06,
+      "loss": 0.3364,
+      "step": 211
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.941361053558564,
+      "learning_rate": 9.948291079656758e-06,
+      "loss": 0.3975,
+      "step": 212
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.494214006153506,
+      "learning_rate": 9.947468492035117e-06,
+      "loss": 0.3528,
+      "step": 213
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.030824926748888,
+      "learning_rate": 9.94663944761328e-06,
+      "loss": 0.3958,
+      "step": 214
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.3471834605784165,
+      "learning_rate": 9.945803947473207e-06,
+      "loss": 0.3171,
+      "step": 215
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.669607529033815,
+      "learning_rate": 9.944961992705288e-06,
+      "loss": 0.2876,
+      "step": 216
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.021312748209139,
+      "learning_rate": 9.94411358440833e-06,
+      "loss": 0.3696,
+      "step": 217
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.327372115755336,
+      "learning_rate": 9.94325872368957e-06,
+      "loss": 0.439,
+      "step": 218
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.211161926706457,
+      "learning_rate": 9.942397411664663e-06,
+      "loss": 0.3384,
+      "step": 219
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.907187518808936,
+      "learning_rate": 9.94152964945768e-06,
+      "loss": 0.2759,
+      "step": 220
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.38762652010209,
+      "learning_rate": 9.940655438201113e-06,
+      "loss": 0.4077,
+      "step": 221
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2311203744807293,
+      "learning_rate": 9.939774779035871e-06,
+      "loss": 0.2517,
+      "step": 222
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5353628982747014,
+      "learning_rate": 9.93888767311128e-06,
+      "loss": 0.2704,
+      "step": 223
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 19.296729355828607,
+      "learning_rate": 9.93799412158507e-06,
+      "loss": 0.4819,
+      "step": 224
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.749679157444635,
+      "learning_rate": 9.937094125623396e-06,
+      "loss": 0.302,
+      "step": 225
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.527590943615032,
+      "learning_rate": 9.936187686400814e-06,
+      "loss": 0.3433,
+      "step": 226
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.810263241562012,
+      "learning_rate": 9.935274805100293e-06,
+      "loss": 0.2896,
+      "step": 227
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.704970511303177,
+      "learning_rate": 9.93435548291321e-06,
+      "loss": 0.3999,
+      "step": 228
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 17.158578443538914,
+      "learning_rate": 9.93342972103934e-06,
+      "loss": 0.2822,
+      "step": 229
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.257483670574077,
+      "learning_rate": 9.932497520686877e-06,
+      "loss": 0.3293,
+      "step": 230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.308958008121345,
+      "learning_rate": 9.931558883072403e-06,
+      "loss": 0.2688,
+      "step": 231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.874907949829154,
+      "learning_rate": 9.930613809420908e-06,
+      "loss": 0.2473,
+      "step": 232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.95285300597983,
+      "learning_rate": 9.929662300965784e-06,
+      "loss": 0.3838,
+      "step": 233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.761092630986912,
+      "learning_rate": 9.928704358948814e-06,
+      "loss": 0.4053,
+      "step": 234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.96369914939653,
+      "learning_rate": 9.927739984620181e-06,
+      "loss": 0.3784,
+      "step": 235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.402307535325615,
+      "learning_rate": 9.926769179238467e-06,
+      "loss": 0.4792,
+      "step": 236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.995921650003602,
+      "learning_rate": 9.925791944070637e-06,
+      "loss": 0.3291,
+      "step": 237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.320489381407661,
+      "learning_rate": 9.924808280392055e-06,
+      "loss": 0.3823,
+      "step": 238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 24.216197995341016,
+      "learning_rate": 9.923818189486472e-06,
+      "loss": 0.4131,
+      "step": 239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 21.177069313976112,
+      "learning_rate": 9.922821672646028e-06,
+      "loss": 0.3452,
+      "step": 240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.660769440538754,
+      "learning_rate": 9.921818731171249e-06,
+      "loss": 0.2998,
+      "step": 241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9074432067654574,
+      "learning_rate": 9.920809366371045e-06,
+      "loss": 0.3103,
+      "step": 242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.713918348847825,
+      "learning_rate": 9.919793579562708e-06,
+      "loss": 0.3053,
+      "step": 243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.214568802834851,
+      "learning_rate": 9.918771372071914e-06,
+      "loss": 0.3252,
+      "step": 244
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.992758675105954,
+      "learning_rate": 9.917742745232718e-06,
+      "loss": 0.5229,
+      "step": 245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 80.73125890235714,
+      "learning_rate": 9.916707700387546e-06,
+      "loss": 0.6064,
+      "step": 246
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 47.73545802394239,
+      "learning_rate": 9.915666238887214e-06,
+      "loss": 0.5312,
+      "step": 247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 457.43546532233944,
+      "learning_rate": 9.914618362090897e-06,
+      "loss": 1.5488,
+      "step": 248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 74.49726591353644,
+      "learning_rate": 9.913564071366152e-06,
+      "loss": 0.5024,
+      "step": 249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 30.044558556244095,
+      "learning_rate": 9.912503368088904e-06,
+      "loss": 0.4414,
+      "step": 250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 72.38577162544867,
+      "learning_rate": 9.911436253643445e-06,
+      "loss": 0.478,
+      "step": 251
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.088481789665817,
+      "learning_rate": 9.910362729422436e-06,
+      "loss": 0.4258,
+      "step": 252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.852521213783266,
+      "learning_rate": 9.909282796826905e-06,
+      "loss": 0.2837,
+      "step": 253
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.21025832564711,
+      "learning_rate": 9.908196457266237e-06,
+      "loss": 0.5088,
+      "step": 254
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.994859910863565,
+      "learning_rate": 9.907103712158189e-06,
+      "loss": 0.3306,
+      "step": 255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.185684791795699,
+      "learning_rate": 9.906004562928865e-06,
+      "loss": 0.3445,
+      "step": 256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.066504963809663,
+      "learning_rate": 9.904899011012737e-06,
+      "loss": 0.5562,
+      "step": 257
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.694929410593049,
+      "learning_rate": 9.90378705785263e-06,
+      "loss": 0.3906,
+      "step": 258
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.404825240573818,
+      "learning_rate": 9.902668704899721e-06,
+      "loss": 0.3984,
+      "step": 259
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 22.062593160791145,
+      "learning_rate": 9.901543953613539e-06,
+      "loss": 0.3345,
+      "step": 260
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.687718086190621,
+      "learning_rate": 9.900412805461968e-06,
+      "loss": 0.3152,
+      "step": 261
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.050038239543859,
+      "learning_rate": 9.899275261921236e-06,
+      "loss": 0.3845,
+      "step": 262
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.994453798099759,
+      "learning_rate": 9.898131324475917e-06,
+      "loss": 0.3428,
+      "step": 263
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.118863370838637,
+      "learning_rate": 9.896980994618933e-06,
+      "loss": 0.3906,
+      "step": 264
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.681355170550944,
+      "learning_rate": 9.895824273851547e-06,
+      "loss": 0.5117,
+      "step": 265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.69584866502579,
+      "learning_rate": 9.894661163683361e-06,
+      "loss": 0.4028,
+      "step": 266
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.076824534590158,
+      "learning_rate": 9.893491665632317e-06,
+      "loss": 0.3638,
+      "step": 267
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.557820141431854,
+      "learning_rate": 9.892315781224695e-06,
+      "loss": 0.3838,
+      "step": 268
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.550290302036172,
+      "learning_rate": 9.891133511995107e-06,
+      "loss": 0.4375,
+      "step": 269
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.17577688245629,
+      "learning_rate": 9.889944859486502e-06,
+      "loss": 0.374,
+      "step": 270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.711507057268035,
+      "learning_rate": 9.888749825250151e-06,
+      "loss": 0.2935,
+      "step": 271
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.632541263610397,
+      "learning_rate": 9.887548410845661e-06,
+      "loss": 0.2922,
+      "step": 272
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.863273490294395,
+      "learning_rate": 9.886340617840968e-06,
+      "loss": 0.4463,
+      "step": 273
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.9387764048235,
+      "learning_rate": 9.885126447812324e-06,
+      "loss": 0.3372,
+      "step": 274
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.594145316261269,
+      "learning_rate": 9.883905902344308e-06,
+      "loss": 0.4131,
+      "step": 275
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.34239364435005,
+      "learning_rate": 9.882678983029819e-06,
+      "loss": 0.4077,
+      "step": 276
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.506485143908037,
+      "learning_rate": 9.881445691470076e-06,
+      "loss": 0.4661,
+      "step": 277
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.469950058998405,
+      "learning_rate": 9.880206029274613e-06,
+      "loss": 0.2883,
+      "step": 278
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.567676614038248,
+      "learning_rate": 9.878959998061274e-06,
+      "loss": 0.4116,
+      "step": 279
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 22.089348374748813,
+      "learning_rate": 9.877707599456224e-06,
+      "loss": 0.3569,
+      "step": 280
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.266117746566283,
+      "learning_rate": 9.87644883509393e-06,
+      "loss": 0.3953,
+      "step": 281
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.750219138788324,
+      "learning_rate": 9.87518370661717e-06,
+      "loss": 0.3555,
+      "step": 282
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.857699870877852,
+      "learning_rate": 9.873912215677026e-06,
+      "loss": 0.4209,
+      "step": 283
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.861545848894709,
+      "learning_rate": 9.872634363932887e-06,
+      "loss": 0.4185,
+      "step": 284
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.30434958719359,
+      "learning_rate": 9.871350153052438e-06,
+      "loss": 0.4937,
+      "step": 285
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.085019111022697,
+      "learning_rate": 9.870059584711668e-06,
+      "loss": 0.3604,
+      "step": 286
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.86813427061735,
+      "learning_rate": 9.868762660594862e-06,
+      "loss": 0.2788,
+      "step": 287
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.927952421579862,
+      "learning_rate": 9.867459382394596e-06,
+      "loss": 0.4551,
+      "step": 288
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.5825058373306415,
+      "learning_rate": 9.866149751811742e-06,
+      "loss": 0.3496,
+      "step": 289
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.959251430631018,
+      "learning_rate": 9.864833770555462e-06,
+      "loss": 0.3433,
+      "step": 290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.225330457084145,
+      "learning_rate": 9.863511440343206e-06,
+      "loss": 0.355,
+      "step": 291
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.32271409405996,
+      "learning_rate": 9.86218276290071e-06,
+      "loss": 0.4282,
+      "step": 292
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.698259184383307,
+      "learning_rate": 9.860847739961992e-06,
+      "loss": 0.4321,
+      "step": 293
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.886673478641496,
+      "learning_rate": 9.859506373269352e-06,
+      "loss": 0.3721,
+      "step": 294
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.45870749719167,
+      "learning_rate": 9.85815866457337e-06,
+      "loss": 0.374,
+      "step": 295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.105670082499662,
+      "learning_rate": 9.856804615632904e-06,
+      "loss": 0.3423,
+      "step": 296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.897513549722222,
+      "learning_rate": 9.855444228215082e-06,
+      "loss": 0.3286,
+      "step": 297
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.232011072698515,
+      "learning_rate": 9.854077504095307e-06,
+      "loss": 0.3979,
+      "step": 298
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.40136589129758,
+      "learning_rate": 9.852704445057256e-06,
+      "loss": 0.3167,
+      "step": 299
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.630281407399583,
+      "learning_rate": 9.851325052892862e-06,
+      "loss": 0.3105,
+      "step": 300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.095257198551451,
+      "learning_rate": 9.849939329402337e-06,
+      "loss": 0.4309,
+      "step": 301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.3124607081094535,
+      "learning_rate": 9.848547276394145e-06,
+      "loss": 0.2393,
+      "step": 302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.6621108715745185,
+      "learning_rate": 9.847148895685018e-06,
+      "loss": 0.2766,
+      "step": 303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.276822965053245,
+      "learning_rate": 9.845744189099938e-06,
+      "loss": 0.3584,
+      "step": 304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.966066811181145,
+      "learning_rate": 9.844333158472153e-06,
+      "loss": 0.3982,
+      "step": 305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.22838280985633,
+      "learning_rate": 9.842915805643156e-06,
+      "loss": 0.4062,
+      "step": 306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.877889990040693,
+      "learning_rate": 9.841492132462696e-06,
+      "loss": 0.3857,
+      "step": 307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.812783198017232,
+      "learning_rate": 9.840062140788765e-06,
+      "loss": 0.3789,
+      "step": 308
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.55666933065705,
+      "learning_rate": 9.838625832487605e-06,
+      "loss": 0.3735,
+      "step": 309
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.482634367427943,
+      "learning_rate": 9.837183209433705e-06,
+      "loss": 0.355,
+      "step": 310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.059988724651665,
+      "learning_rate": 9.835734273509787e-06,
+      "loss": 0.3679,
+      "step": 311
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.989003274898286,
+      "learning_rate": 9.834279026606817e-06,
+      "loss": 0.3071,
+      "step": 312
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.364610361376657,
+      "learning_rate": 9.832817470623996e-06,
+      "loss": 0.2847,
+      "step": 313
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.285864688017158,
+      "learning_rate": 9.831349607468763e-06,
+      "loss": 0.4243,
+      "step": 314
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.842158305094546,
+      "learning_rate": 9.829875439056778e-06,
+      "loss": 0.499,
+      "step": 315
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.102583021102224,
+      "learning_rate": 9.82839496731194e-06,
+      "loss": 0.3315,
+      "step": 316
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.270662490891907,
+      "learning_rate": 9.82690819416637e-06,
+      "loss": 0.3857,
+      "step": 317
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 14.207899360475471,
+      "learning_rate": 9.825415121560414e-06,
+      "loss": 0.3691,
+      "step": 318
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.563526640000224,
+      "learning_rate": 9.823915751442637e-06,
+      "loss": 0.4019,
+      "step": 319
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.102100149464281,
+      "learning_rate": 9.822410085769824e-06,
+      "loss": 0.4448,
+      "step": 320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.170764530861115,
+      "learning_rate": 9.820898126506978e-06,
+      "loss": 0.4355,
+      "step": 321
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.361532547818666,
+      "learning_rate": 9.819379875627315e-06,
+      "loss": 0.5737,
+      "step": 322
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.207584899061887,
+      "learning_rate": 9.817855335112256e-06,
+      "loss": 0.3062,
+      "step": 323
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6306970023547236,
+      "learning_rate": 9.81632450695144e-06,
+      "loss": 0.3813,
+      "step": 324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.043471140562569,
+      "learning_rate": 9.814787393142706e-06,
+      "loss": 0.3486,
+      "step": 325
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.318232139731164,
+      "learning_rate": 9.813243995692097e-06,
+      "loss": 0.3604,
+      "step": 326
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.02908193075617,
+      "learning_rate": 9.811694316613858e-06,
+      "loss": 0.2979,
+      "step": 327
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.238986428134632,
+      "learning_rate": 9.81013835793043e-06,
+      "loss": 0.3198,
+      "step": 328
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.321869024186741,
+      "learning_rate": 9.808576121672451e-06,
+      "loss": 0.3918,
+      "step": 329
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.246292854815053,
+      "learning_rate": 9.80700760987875e-06,
+      "loss": 0.3896,
+      "step": 330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.879465757122626,
+      "learning_rate": 9.805432824596347e-06,
+      "loss": 0.4209,
+      "step": 331
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.632762912574986,
+      "learning_rate": 9.803851767880453e-06,
+      "loss": 0.46,
+      "step": 332
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.166213429534091,
+      "learning_rate": 9.802264441794456e-06,
+      "loss": 0.3696,
+      "step": 333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.028787655662285,
+      "learning_rate": 9.80067084840993e-06,
+      "loss": 0.3848,
+      "step": 334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.409530450356453,
+      "learning_rate": 9.799070989806632e-06,
+      "loss": 0.4238,
+      "step": 335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.550791079040542,
+      "learning_rate": 9.797464868072489e-06,
+      "loss": 0.4736,
+      "step": 336
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.208644714601721,
+      "learning_rate": 9.795852485303603e-06,
+      "loss": 0.3467,
+      "step": 337
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.031444109956205,
+      "learning_rate": 9.794233843604253e-06,
+      "loss": 0.3408,
+      "step": 338
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.526063654369636,
+      "learning_rate": 9.79260894508688e-06,
+      "loss": 0.4165,
+      "step": 339
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.382980378228476,
+      "learning_rate": 9.790977791872094e-06,
+      "loss": 0.3484,
+      "step": 340
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.535557300305796,
+      "learning_rate": 9.789340386088663e-06,
+      "loss": 0.226,
+      "step": 341
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.835319385135228,
+      "learning_rate": 9.787696729873524e-06,
+      "loss": 0.2993,
+      "step": 342
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.780540303622722,
+      "learning_rate": 9.786046825371762e-06,
+      "loss": 0.3242,
+      "step": 343
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.777196770510356,
+      "learning_rate": 9.784390674736622e-06,
+      "loss": 0.3191,
+      "step": 344
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.41929125161025,
+      "learning_rate": 9.782728280129498e-06,
+      "loss": 0.4746,
+      "step": 345
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.702678178619091,
+      "learning_rate": 9.781059643719937e-06,
+      "loss": 0.2954,
+      "step": 346
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9976029568270173,
+      "learning_rate": 9.779384767685624e-06,
+      "loss": 0.3545,
+      "step": 347
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.4051056755231075,
+      "learning_rate": 9.777703654212393e-06,
+      "loss": 0.3484,
+      "step": 348
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.560685825784729,
+      "learning_rate": 9.776016305494221e-06,
+      "loss": 0.3101,
+      "step": 349
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 14.91487312816826,
+      "learning_rate": 9.774322723733216e-06,
+      "loss": 0.3208,
+      "step": 350
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.192935152750769,
+      "learning_rate": 9.772622911139622e-06,
+      "loss": 0.3516,
+      "step": 351
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9491468915262495,
+      "learning_rate": 9.77091686993182e-06,
+      "loss": 0.3364,
+      "step": 352
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.7977429964779645,
+      "learning_rate": 9.769204602336312e-06,
+      "loss": 0.3745,
+      "step": 353
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.1443798831828795,
+      "learning_rate": 9.767486110587731e-06,
+      "loss": 0.231,
+      "step": 354
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.050024917861675,
+      "learning_rate": 9.765761396928832e-06,
+      "loss": 0.4551,
+      "step": 355
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.271107303864191,
+      "learning_rate": 9.76403046361049e-06,
+      "loss": 0.3251,
+      "step": 356
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.844633968924002,
+      "learning_rate": 9.762293312891694e-06,
+      "loss": 0.3667,
+      "step": 357
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.020180923316104,
+      "learning_rate": 9.760549947039556e-06,
+      "loss": 0.3711,
+      "step": 358
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5204599658674267,
+      "learning_rate": 9.758800368329286e-06,
+      "loss": 0.374,
+      "step": 359
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.632594415244508,
+      "learning_rate": 9.757044579044215e-06,
+      "loss": 0.283,
+      "step": 360
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.691576618895846,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.3374,
+      "step": 361
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.90514504124684,
+      "learning_rate": 9.753514377923483e-06,
+      "loss": 0.4585,
+      "step": 362
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.889761216007317,
+      "learning_rate": 9.75173997069499e-06,
+      "loss": 0.3657,
+      "step": 363
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.608948824382592,
+      "learning_rate": 9.749959362106017e-06,
+      "loss": 0.3335,
+      "step": 364
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.321389479449753,
+      "learning_rate": 9.748172554480386e-06,
+      "loss": 0.3354,
+      "step": 365
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.7975560133662025,
+      "learning_rate": 9.74637955015001e-06,
+      "loss": 0.373,
+      "step": 366
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4517928857606126,
+      "learning_rate": 9.744580351454885e-06,
+      "loss": 0.356,
+      "step": 367
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.970461444223229,
+      "learning_rate": 9.742774960743095e-06,
+      "loss": 0.4697,
+      "step": 368
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0042255031585547,
+      "learning_rate": 9.740963380370805e-06,
+      "loss": 0.3162,
+      "step": 369
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3542088921489794,
+      "learning_rate": 9.739145612702256e-06,
+      "loss": 0.3652,
+      "step": 370
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.562074967590413,
+      "learning_rate": 9.737321660109767e-06,
+      "loss": 0.2742,
+      "step": 371
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.704419909505527,
+      "learning_rate": 9.735491524973723e-06,
+      "loss": 0.3447,
+      "step": 372
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4440854639398015,
+      "learning_rate": 9.733655209682582e-06,
+      "loss": 0.2937,
+      "step": 373
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.944376267335289,
+      "learning_rate": 9.73181271663287e-06,
+      "loss": 0.3145,
+      "step": 374
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.761740983434704,
+      "learning_rate": 9.72996404822917e-06,
+      "loss": 0.314,
+      "step": 375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.123834596071273,
+      "learning_rate": 9.728109206884125e-06,
+      "loss": 0.3877,
+      "step": 376
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.126888262305346,
+      "learning_rate": 9.726248195018438e-06,
+      "loss": 0.3853,
+      "step": 377
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.739885809786603,
+      "learning_rate": 9.724381015060862e-06,
+      "loss": 0.3228,
+      "step": 378
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.163198823540041,
+      "learning_rate": 9.7225076694482e-06,
+      "loss": 0.417,
+      "step": 379
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.72701396718307,
+      "learning_rate": 9.720628160625302e-06,
+      "loss": 0.4673,
+      "step": 380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.272438799873677,
+      "learning_rate": 9.718742491045061e-06,
+      "loss": 0.356,
+      "step": 381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.759964610687544,
+      "learning_rate": 9.716850663168412e-06,
+      "loss": 0.3999,
+      "step": 382
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.06157403761296,
+      "learning_rate": 9.714952679464324e-06,
+      "loss": 0.3618,
+      "step": 383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.045322402057133,
+      "learning_rate": 9.713048542409802e-06,
+      "loss": 0.2996,
+      "step": 384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.029567617527702,
+      "learning_rate": 9.711138254489882e-06,
+      "loss": 0.3413,
+      "step": 385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.01191168607786,
+      "learning_rate": 9.709221818197626e-06,
+      "loss": 0.3386,
+      "step": 386
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.946016414971407,
+      "learning_rate": 9.707299236034118e-06,
+      "loss": 0.4673,
+      "step": 387
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.248794574604268,
+      "learning_rate": 9.705370510508468e-06,
+      "loss": 0.4399,
+      "step": 388
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.678983818928527,
+      "learning_rate": 9.7034356441378e-06,
+      "loss": 0.4497,
+      "step": 389
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.891343188013226,
+      "learning_rate": 9.701494639447254e-06,
+      "loss": 0.3381,
+      "step": 390
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.169739129534069,
+      "learning_rate": 9.699547498969978e-06,
+      "loss": 0.3506,
+      "step": 391
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.828215430805521,
+      "learning_rate": 9.697594225247132e-06,
+      "loss": 0.3325,
+      "step": 392
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.185524654640735,
+      "learning_rate": 9.695634820827874e-06,
+      "loss": 0.4868,
+      "step": 393
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.883461777913633,
+      "learning_rate": 9.693669288269371e-06,
+      "loss": 0.2725,
+      "step": 394
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.865872859094069,
+      "learning_rate": 9.691697630136782e-06,
+      "loss": 0.501,
+      "step": 395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.66085971880933,
+      "learning_rate": 9.689719849003261e-06,
+      "loss": 0.3564,
+      "step": 396
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.954104458976826,
+      "learning_rate": 9.687735947449955e-06,
+      "loss": 0.3887,
+      "step": 397
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.909698727184145,
+      "learning_rate": 9.685745928065997e-06,
+      "loss": 0.4189,
+      "step": 398
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.10032281855974,
+      "learning_rate": 9.683749793448502e-06,
+      "loss": 0.4219,
+      "step": 399
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.807429325888113,
+      "learning_rate": 9.681747546202573e-06,
+      "loss": 0.3638,
+      "step": 400
+    },
+    {
+      "epoch": 0.56,
+      "eval_avg_AUC": 0.739011770033394,
+      "eval_avg_Accuracy": 0.6405421087533156,
+      "eval_avg_Accuracy-right": 0.9936089735228903,
+      "eval_avg_Accuracy-wrong": 0.02490334318853764,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6355598729823532,
+      "eval_last_AUC": 0.7519614905451947,
+      "eval_last_Accuracy": 0.707145225464191,
+      "eval_last_Accuracy-right": 0.8591365592800313,
+      "eval_last_Accuracy-wrong": 0.44211962701842167,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6316826783375092,
+      "eval_max_AUC": 0.6214924965885089,
+      "eval_max_Accuracy": 0.6354442970822282,
+      "eval_max_Accuracy-right": 0.9991522107734446,
+      "eval_max_Accuracy-wrong": 0.0012508528542187855,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5435114003611005,
+      "eval_min_AUC": 0.7523072634005679,
+      "eval_min_Accuracy": 0.7101293103448276,
+      "eval_min_Accuracy-right": 0.8532672492500326,
+      "eval_min_Accuracy-wrong": 0.4605412781441892,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6298057073500145,
+      "eval_prod_AUC": 0.753607757628123,
+      "eval_prod_Accuracy": 0.6732841511936339,
+      "eval_prod_Accuracy-right": 0.6106691013434199,
+      "eval_prod_Accuracy-wrong": 0.7824653172617694,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6312054794672889,
+      "eval_runtime": 668.5898,
+      "eval_samples_per_second": 36.088,
+      "eval_steps_per_second": 1.128,
+      "eval_sum_AUC": 0.5354536775091712,
+      "eval_sum_Accuracy": 0.6356100795755968,
+      "eval_sum_Accuracy-right": 0.9999347854441111,
+      "eval_sum_Accuracy-wrong": 0.0003411416875142142,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.59198922296681,
+      "step": 400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.0973395577220435,
+      "learning_rate": 9.679739188941283e-06,
+      "loss": 0.3284,
+      "step": 401
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.72434985670905,
+      "learning_rate": 9.677724724285686e-06,
+      "loss": 0.231,
+      "step": 402
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.867375323282493,
+      "learning_rate": 9.675704154864796e-06,
+      "loss": 0.249,
+      "step": 403
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.350466604317953,
+      "learning_rate": 9.673677483315604e-06,
+      "loss": 0.4512,
+      "step": 404
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.493491608018028,
+      "learning_rate": 9.671644712283061e-06,
+      "loss": 0.397,
+      "step": 405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8103562587262036,
+      "learning_rate": 9.66960584442008e-06,
+      "loss": 0.3433,
+      "step": 406
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.982812255506099,
+      "learning_rate": 9.667560882387526e-06,
+      "loss": 0.5142,
+      "step": 407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.234668733676103,
+      "learning_rate": 9.665509828854223e-06,
+      "loss": 0.4282,
+      "step": 408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.488234670630249,
+      "learning_rate": 9.663452686496943e-06,
+      "loss": 0.3245,
+      "step": 409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.731846597765078,
+      "learning_rate": 9.661389458000402e-06,
+      "loss": 0.416,
+      "step": 410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.543648117911364,
+      "learning_rate": 9.659320146057263e-06,
+      "loss": 0.3525,
+      "step": 411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.608800604488739,
+      "learning_rate": 9.657244753368124e-06,
+      "loss": 0.3828,
+      "step": 412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.697652256026757,
+      "learning_rate": 9.655163282641522e-06,
+      "loss": 0.3643,
+      "step": 413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.4259794202270895,
+      "learning_rate": 9.653075736593924e-06,
+      "loss": 0.3899,
+      "step": 414
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.601437494098513,
+      "learning_rate": 9.65098211794973e-06,
+      "loss": 0.3474,
+      "step": 415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.790410586549175,
+      "learning_rate": 9.648882429441258e-06,
+      "loss": 0.439,
+      "step": 416
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.404505453488826,
+      "learning_rate": 9.646776673808753e-06,
+      "loss": 0.2505,
+      "step": 417
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.336181893315071,
+      "learning_rate": 9.644664853800379e-06,
+      "loss": 0.4363,
+      "step": 418
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.876461982040368,
+      "learning_rate": 9.64254697217221e-06,
+      "loss": 0.3208,
+      "step": 419
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.0887862168716,
+      "learning_rate": 9.640423031688232e-06,
+      "loss": 0.4302,
+      "step": 420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.260177144126579,
+      "learning_rate": 9.638293035120342e-06,
+      "loss": 0.3252,
+      "step": 421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.148234079819579,
+      "learning_rate": 9.636156985248337e-06,
+      "loss": 0.3071,
+      "step": 422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.986753666695353,
+      "learning_rate": 9.634014884859912e-06,
+      "loss": 0.3579,
+      "step": 423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.118232367633379,
+      "learning_rate": 9.631866736750664e-06,
+      "loss": 0.4697,
+      "step": 424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.20418924907706,
+      "learning_rate": 9.62971254372408e-06,
+      "loss": 0.3264,
+      "step": 425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.690084928197341,
+      "learning_rate": 9.627552308591534e-06,
+      "loss": 0.3586,
+      "step": 426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.7950909887199495,
+      "learning_rate": 9.62538603417229e-06,
+      "loss": 0.3379,
+      "step": 427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.5405819671009775,
+      "learning_rate": 9.62321372329349e-06,
+      "loss": 0.2805,
+      "step": 428
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.846928097188302,
+      "learning_rate": 9.621035378790155e-06,
+      "loss": 0.4617,
+      "step": 429
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.848871910287274,
+      "learning_rate": 9.618851003505178e-06,
+      "loss": 0.3716,
+      "step": 430
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.148980265513956,
+      "learning_rate": 9.616660600289329e-06,
+      "loss": 0.3521,
+      "step": 431
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.811947692206455,
+      "learning_rate": 9.614464172001238e-06,
+      "loss": 0.3501,
+      "step": 432
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9933332730504993,
+      "learning_rate": 9.612261721507402e-06,
+      "loss": 0.2979,
+      "step": 433
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8745769443179983,
+      "learning_rate": 9.610053251682177e-06,
+      "loss": 0.3584,
+      "step": 434
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.266715529851013,
+      "learning_rate": 9.60783876540777e-06,
+      "loss": 0.2698,
+      "step": 435
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.656400213117868,
+      "learning_rate": 9.60561826557425e-06,
+      "loss": 0.2561,
+      "step": 436
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.2277138842817585,
+      "learning_rate": 9.603391755079527e-06,
+      "loss": 0.2891,
+      "step": 437
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.718379263623455,
+      "learning_rate": 9.601159236829353e-06,
+      "loss": 0.2439,
+      "step": 438
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.677263189743773,
+      "learning_rate": 9.598920713737326e-06,
+      "loss": 0.2922,
+      "step": 439
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.274966341855977,
+      "learning_rate": 9.596676188724878e-06,
+      "loss": 0.457,
+      "step": 440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.735856283995487,
+      "learning_rate": 9.594425664721275e-06,
+      "loss": 0.4072,
+      "step": 441
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.758587151558096,
+      "learning_rate": 9.592169144663611e-06,
+      "loss": 0.4785,
+      "step": 442
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.151118526547789,
+      "learning_rate": 9.589906631496809e-06,
+      "loss": 0.479,
+      "step": 443
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.7644402904286105,
+      "learning_rate": 9.587638128173608e-06,
+      "loss": 0.3743,
+      "step": 444
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 15.508316167131332,
+      "learning_rate": 9.585363637654567e-06,
+      "loss": 0.3662,
+      "step": 445
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.650635405510034,
+      "learning_rate": 9.58308316290806e-06,
+      "loss": 0.3906,
+      "step": 446
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.906492411056538,
+      "learning_rate": 9.580796706910268e-06,
+      "loss": 0.3877,
+      "step": 447
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.991812063882783,
+      "learning_rate": 9.57850427264518e-06,
+      "loss": 0.3237,
+      "step": 448
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.1839726525461725,
+      "learning_rate": 9.576205863104588e-06,
+      "loss": 0.3889,
+      "step": 449
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.368165359545814,
+      "learning_rate": 9.57390148128808e-06,
+      "loss": 0.3711,
+      "step": 450
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.765825074126903,
+      "learning_rate": 9.57159113020304e-06,
+      "loss": 0.436,
+      "step": 451
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.087459022782288,
+      "learning_rate": 9.56927481286464e-06,
+      "loss": 0.2715,
+      "step": 452
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.0154058410940205,
+      "learning_rate": 9.566952532295838e-06,
+      "loss": 0.3369,
+      "step": 453
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.251652300039112,
+      "learning_rate": 9.56462429152738e-06,
+      "loss": 0.3291,
+      "step": 454
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.266917054842654,
+      "learning_rate": 9.562290093597787e-06,
+      "loss": 0.3792,
+      "step": 455
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.12922823136087,
+      "learning_rate": 9.559949941553351e-06,
+      "loss": 0.4019,
+      "step": 456
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4247909263173963,
+      "learning_rate": 9.55760383844814e-06,
+      "loss": 0.301,
+      "step": 457
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.508594689462784,
+      "learning_rate": 9.555251787343981e-06,
+      "loss": 0.3604,
+      "step": 458
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.3287604729630615,
+      "learning_rate": 9.552893791310479e-06,
+      "loss": 0.3472,
+      "step": 459
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 21.863726890939635,
+      "learning_rate": 9.550529853424979e-06,
+      "loss": 0.2971,
+      "step": 460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.403657442836438,
+      "learning_rate": 9.548159976772593e-06,
+      "loss": 0.394,
+      "step": 461
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4810484711051353,
+      "learning_rate": 9.545784164446177e-06,
+      "loss": 0.4043,
+      "step": 462
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.6618922850299205,
+      "learning_rate": 9.54340241954634e-06,
+      "loss": 0.4141,
+      "step": 463
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.38130169156453,
+      "learning_rate": 9.541014745181422e-06,
+      "loss": 0.436,
+      "step": 464
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.435300927095303,
+      "learning_rate": 9.538621144467518e-06,
+      "loss": 0.4092,
+      "step": 465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.0920548258628315,
+      "learning_rate": 9.536221620528442e-06,
+      "loss": 0.3691,
+      "step": 466
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2035942067050005,
+      "learning_rate": 9.533816176495746e-06,
+      "loss": 0.3059,
+      "step": 467
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.096815279310491,
+      "learning_rate": 9.531404815508706e-06,
+      "loss": 0.293,
+      "step": 468
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9150384613811937,
+      "learning_rate": 9.528987540714323e-06,
+      "loss": 0.3677,
+      "step": 469
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.928145383497267,
+      "learning_rate": 9.52656435526731e-06,
+      "loss": 0.3545,
+      "step": 470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.203135956003578,
+      "learning_rate": 9.524135262330098e-06,
+      "loss": 0.2915,
+      "step": 471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3132006539032917,
+      "learning_rate": 9.521700265072829e-06,
+      "loss": 0.4077,
+      "step": 472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.997224913584175,
+      "learning_rate": 9.519259366673348e-06,
+      "loss": 0.3027,
+      "step": 473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.384616980617786,
+      "learning_rate": 9.5168125703172e-06,
+      "loss": 0.3374,
+      "step": 474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.96704646176088,
+      "learning_rate": 9.51435987919763e-06,
+      "loss": 0.3765,
+      "step": 475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.874701133594703,
+      "learning_rate": 9.511901296515578e-06,
+      "loss": 0.4399,
+      "step": 476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.968611418639268,
+      "learning_rate": 9.509436825479666e-06,
+      "loss": 0.3774,
+      "step": 477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.371088587578784,
+      "learning_rate": 9.506966469306208e-06,
+      "loss": 0.3638,
+      "step": 478
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.609554797766558,
+      "learning_rate": 9.504490231219194e-06,
+      "loss": 0.3066,
+      "step": 479
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.689872466930098,
+      "learning_rate": 9.502008114450294e-06,
+      "loss": 0.3638,
+      "step": 480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.350258225367147,
+      "learning_rate": 9.499520122238846e-06,
+      "loss": 0.2571,
+      "step": 481
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.094728759060189,
+      "learning_rate": 9.497026257831856e-06,
+      "loss": 0.458,
+      "step": 482
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.775719543208952,
+      "learning_rate": 9.494526524483998e-06,
+      "loss": 0.397,
+      "step": 483
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5606720950981066,
+      "learning_rate": 9.492020925457603e-06,
+      "loss": 0.2744,
+      "step": 484
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.573916945961893,
+      "learning_rate": 9.489509464022656e-06,
+      "loss": 0.2773,
+      "step": 485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.70869909668727,
+      "learning_rate": 9.486992143456792e-06,
+      "loss": 0.4873,
+      "step": 486
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.196660143237354,
+      "learning_rate": 9.484468967045298e-06,
+      "loss": 0.3779,
+      "step": 487
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.884190064857444,
+      "learning_rate": 9.481939938081098e-06,
+      "loss": 0.4043,
+      "step": 488
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.934492447097597,
+      "learning_rate": 9.479405059864755e-06,
+      "loss": 0.3755,
+      "step": 489
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.043831713100799,
+      "learning_rate": 9.476864335704467e-06,
+      "loss": 0.2988,
+      "step": 490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.612532059881081,
+      "learning_rate": 9.47431776891606e-06,
+      "loss": 0.3506,
+      "step": 491
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.582483175248363,
+      "learning_rate": 9.471765362822987e-06,
+      "loss": 0.3843,
+      "step": 492
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6860805574498037,
+      "learning_rate": 9.46920712075632e-06,
+      "loss": 0.3721,
+      "step": 493
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.221988476022411,
+      "learning_rate": 9.466643046054746e-06,
+      "loss": 0.3977,
+      "step": 494
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.489956352950597,
+      "learning_rate": 9.46407314206457e-06,
+      "loss": 0.2979,
+      "step": 495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8985660749115874,
+      "learning_rate": 9.461497412139697e-06,
+      "loss": 0.377,
+      "step": 496
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.250572377308685,
+      "learning_rate": 9.45891585964164e-06,
+      "loss": 0.3257,
+      "step": 497
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 22.452499865903523,
+      "learning_rate": 9.45632848793951e-06,
+      "loss": 0.5527,
+      "step": 498
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3538637239150733,
+      "learning_rate": 9.45373530041001e-06,
+      "loss": 0.3008,
+      "step": 499
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.156035271973316,
+      "learning_rate": 9.451136300437438e-06,
+      "loss": 0.3154,
+      "step": 500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7641333975464026,
+      "learning_rate": 9.448531491413673e-06,
+      "loss": 0.2654,
+      "step": 501
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.314852750679917,
+      "learning_rate": 9.445920876738178e-06,
+      "loss": 0.4546,
+      "step": 502
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.13006926864923,
+      "learning_rate": 9.44330445981799e-06,
+      "loss": 0.3599,
+      "step": 503
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.817648341213029,
+      "learning_rate": 9.440682244067724e-06,
+      "loss": 0.3496,
+      "step": 504
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.687869913462713,
+      "learning_rate": 9.438054232909555e-06,
+      "loss": 0.2935,
+      "step": 505
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.216591645483212,
+      "learning_rate": 9.435420429773227e-06,
+      "loss": 0.3218,
+      "step": 506
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.020794702831074,
+      "learning_rate": 9.432780838096042e-06,
+      "loss": 0.2712,
+      "step": 507
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6993505355325706,
+      "learning_rate": 9.430135461322858e-06,
+      "loss": 0.2871,
+      "step": 508
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.845669939852009,
+      "learning_rate": 9.427484302906075e-06,
+      "loss": 0.4927,
+      "step": 509
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.693673254880407,
+      "learning_rate": 9.42482736630565e-06,
+      "loss": 0.3738,
+      "step": 510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.617862747787442,
+      "learning_rate": 9.422164654989073e-06,
+      "loss": 0.2507,
+      "step": 511
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.309591890681132,
+      "learning_rate": 9.419496172431372e-06,
+      "loss": 0.3662,
+      "step": 512
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.92948694789209,
+      "learning_rate": 9.416821922115108e-06,
+      "loss": 0.3384,
+      "step": 513
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.33629303616592,
+      "learning_rate": 9.41414190753037e-06,
+      "loss": 0.4226,
+      "step": 514
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.865366832993308,
+      "learning_rate": 9.411456132174768e-06,
+      "loss": 0.3638,
+      "step": 515
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.99512996180443,
+      "learning_rate": 9.408764599553429e-06,
+      "loss": 0.2883,
+      "step": 516
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.34499186208262,
+      "learning_rate": 9.406067313179e-06,
+      "loss": 0.3494,
+      "step": 517
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.987978630236364,
+      "learning_rate": 9.403364276571628e-06,
+      "loss": 0.3223,
+      "step": 518
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.7719266480428875,
+      "learning_rate": 9.400655493258973e-06,
+      "loss": 0.3784,
+      "step": 519
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.856671055135882,
+      "learning_rate": 9.397940966776187e-06,
+      "loss": 0.4038,
+      "step": 520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.948829587068962,
+      "learning_rate": 9.395220700665924e-06,
+      "loss": 0.3945,
+      "step": 521
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.750076935862083,
+      "learning_rate": 9.392494698478326e-06,
+      "loss": 0.3638,
+      "step": 522
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.313798031062023,
+      "learning_rate": 9.389762963771018e-06,
+      "loss": 0.3208,
+      "step": 523
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.35386174026813,
+      "learning_rate": 9.387025500109113e-06,
+      "loss": 0.2539,
+      "step": 524
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.295949282704482,
+      "learning_rate": 9.384282311065194e-06,
+      "loss": 0.2852,
+      "step": 525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.2667524906840715,
+      "learning_rate": 9.381533400219319e-06,
+      "loss": 0.333,
+      "step": 526
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4598908344295887,
+      "learning_rate": 9.378778771159014e-06,
+      "loss": 0.2588,
+      "step": 527
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.582033884112395,
+      "learning_rate": 9.376018427479267e-06,
+      "loss": 0.3342,
+      "step": 528
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.950964519008354,
+      "learning_rate": 9.373252372782525e-06,
+      "loss": 0.3726,
+      "step": 529
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.362695431063013,
+      "learning_rate": 9.370480610678684e-06,
+      "loss": 0.2782,
+      "step": 530
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.442064989888123,
+      "learning_rate": 9.367703144785097e-06,
+      "loss": 0.2595,
+      "step": 531
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3291389075667164,
+      "learning_rate": 9.36491997872655e-06,
+      "loss": 0.2615,
+      "step": 532
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.2739299287910075,
+      "learning_rate": 9.362131116135276e-06,
+      "loss": 0.3774,
+      "step": 533
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.604584004549958,
+      "learning_rate": 9.35933656065094e-06,
+      "loss": 0.2505,
+      "step": 534
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.063865180249807,
+      "learning_rate": 9.356536315920637e-06,
+      "loss": 0.3872,
+      "step": 535
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.063817877813009,
+      "learning_rate": 9.353730385598887e-06,
+      "loss": 0.4224,
+      "step": 536
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.894215731698285,
+      "learning_rate": 9.35091877334763e-06,
+      "loss": 0.3311,
+      "step": 537
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.463734806193173,
+      "learning_rate": 9.348101482836218e-06,
+      "loss": 0.3896,
+      "step": 538
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.653639386815119,
+      "learning_rate": 9.345278517741422e-06,
+      "loss": 0.2737,
+      "step": 539
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.44242641618282,
+      "learning_rate": 9.342449881747408e-06,
+      "loss": 0.2944,
+      "step": 540
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.094121429343713,
+      "learning_rate": 9.339615578545753e-06,
+      "loss": 0.3721,
+      "step": 541
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.320399182768416,
+      "learning_rate": 9.336775611835422e-06,
+      "loss": 0.4048,
+      "step": 542
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.092201198887879,
+      "learning_rate": 9.333929985322773e-06,
+      "loss": 0.2407,
+      "step": 543
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.507236130199413,
+      "learning_rate": 9.331078702721558e-06,
+      "loss": 0.3586,
+      "step": 544
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.181279708222159,
+      "learning_rate": 9.3282217677529e-06,
+      "loss": 0.3555,
+      "step": 545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.880357882444246,
+      "learning_rate": 9.325359184145307e-06,
+      "loss": 0.2969,
+      "step": 546
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.294816682923946,
+      "learning_rate": 9.322490955634652e-06,
+      "loss": 0.4497,
+      "step": 547
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.701911937528973,
+      "learning_rate": 9.319617085964177e-06,
+      "loss": 0.4434,
+      "step": 548
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.676739180016138,
+      "learning_rate": 9.316737578884493e-06,
+      "loss": 0.4478,
+      "step": 549
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.9356540338749255,
+      "learning_rate": 9.313852438153558e-06,
+      "loss": 0.3389,
+      "step": 550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.04972991858034,
+      "learning_rate": 9.310961667536689e-06,
+      "loss": 0.3848,
+      "step": 551
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.108678160636451,
+      "learning_rate": 9.308065270806548e-06,
+      "loss": 0.4185,
+      "step": 552
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.689712954770458,
+      "learning_rate": 9.30516325174314e-06,
+      "loss": 0.3582,
+      "step": 553
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.8076252610464,
+      "learning_rate": 9.302255614133807e-06,
+      "loss": 0.3589,
+      "step": 554
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.228325356435034,
+      "learning_rate": 9.299342361773224e-06,
+      "loss": 0.3179,
+      "step": 555
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4956186646835157,
+      "learning_rate": 9.296423498463396e-06,
+      "loss": 0.2749,
+      "step": 556
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0389620334473846,
+      "learning_rate": 9.293499028013643e-06,
+      "loss": 0.2407,
+      "step": 557
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.314037854757126,
+      "learning_rate": 9.290568954240614e-06,
+      "loss": 0.3914,
+      "step": 558
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.966533659465946,
+      "learning_rate": 9.287633280968263e-06,
+      "loss": 0.3093,
+      "step": 559
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.77899486348961,
+      "learning_rate": 9.28469201202785e-06,
+      "loss": 0.2808,
+      "step": 560
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.05563123020932,
+      "learning_rate": 9.281745151257946e-06,
+      "loss": 0.3191,
+      "step": 561
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.62943978048517,
+      "learning_rate": 9.278792702504412e-06,
+      "loss": 0.4077,
+      "step": 562
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.352023733514466,
+      "learning_rate": 9.275834669620407e-06,
+      "loss": 0.3813,
+      "step": 563
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.807008710964796,
+      "learning_rate": 9.272871056466372e-06,
+      "loss": 0.3281,
+      "step": 564
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.639521106501943,
+      "learning_rate": 9.26990186691004e-06,
+      "loss": 0.4429,
+      "step": 565
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.527982723543461,
+      "learning_rate": 9.26692710482641e-06,
+      "loss": 0.4062,
+      "step": 566
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.882949504767341,
+      "learning_rate": 9.263946774097761e-06,
+      "loss": 0.3613,
+      "step": 567
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.51468392637906,
+      "learning_rate": 9.26096087861364e-06,
+      "loss": 0.262,
+      "step": 568
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.3222997593191,
+      "learning_rate": 9.257969422270851e-06,
+      "loss": 0.3496,
+      "step": 569
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.243942321003618,
+      "learning_rate": 9.25497240897346e-06,
+      "loss": 0.2915,
+      "step": 570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.546160433660787,
+      "learning_rate": 9.251969842632785e-06,
+      "loss": 0.3118,
+      "step": 571
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.264926964871067,
+      "learning_rate": 9.248961727167384e-06,
+      "loss": 0.3462,
+      "step": 572
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.2213435910749935,
+      "learning_rate": 9.245948066503071e-06,
+      "loss": 0.3674,
+      "step": 573
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.676417224017262,
+      "learning_rate": 9.242928864572885e-06,
+      "loss": 0.3262,
+      "step": 574
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.21479826662598,
+      "learning_rate": 9.239904125317099e-06,
+      "loss": 0.3469,
+      "step": 575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.458206623735813,
+      "learning_rate": 9.236873852683213e-06,
+      "loss": 0.2759,
+      "step": 576
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.132934677399477,
+      "learning_rate": 9.233838050625953e-06,
+      "loss": 0.2644,
+      "step": 577
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.532534038298481,
+      "learning_rate": 9.230796723107259e-06,
+      "loss": 0.2756,
+      "step": 578
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 43.879112882130826,
+      "learning_rate": 9.227749874096277e-06,
+      "loss": 0.4778,
+      "step": 579
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.075397213893753,
+      "learning_rate": 9.224697507569364e-06,
+      "loss": 0.2522,
+      "step": 580
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.669724484205899,
+      "learning_rate": 9.221639627510076e-06,
+      "loss": 0.2422,
+      "step": 581
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.37027744809432,
+      "learning_rate": 9.218576237909167e-06,
+      "loss": 0.3809,
+      "step": 582
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.301839921986024,
+      "learning_rate": 9.215507342764578e-06,
+      "loss": 0.3716,
+      "step": 583
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.590444639809245,
+      "learning_rate": 9.212432946081437e-06,
+      "loss": 0.3037,
+      "step": 584
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1550114958916438,
+      "learning_rate": 9.209353051872054e-06,
+      "loss": 0.3247,
+      "step": 585
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.747367863620667,
+      "learning_rate": 9.206267664155906e-06,
+      "loss": 0.3618,
+      "step": 586
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3303176765734253,
+      "learning_rate": 9.203176786959649e-06,
+      "loss": 0.334,
+      "step": 587
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6574223241252932,
+      "learning_rate": 9.200080424317096e-06,
+      "loss": 0.2776,
+      "step": 588
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.345220746316483,
+      "learning_rate": 9.196978580269226e-06,
+      "loss": 0.3699,
+      "step": 589
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.479341376438608,
+      "learning_rate": 9.193871258864162e-06,
+      "loss": 0.4028,
+      "step": 590
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.574903822171784,
+      "learning_rate": 9.190758464157184e-06,
+      "loss": 0.2563,
+      "step": 591
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5870901846611516,
+      "learning_rate": 9.18764020021071e-06,
+      "loss": 0.3425,
+      "step": 592
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.608041862032128,
+      "learning_rate": 9.184516471094298e-06,
+      "loss": 0.5166,
+      "step": 593
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0382776775826055,
+      "learning_rate": 9.181387280884637e-06,
+      "loss": 0.3872,
+      "step": 594
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.9513191106169065,
+      "learning_rate": 9.178252633665548e-06,
+      "loss": 0.376,
+      "step": 595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.02684226166284,
+      "learning_rate": 9.175112533527963e-06,
+      "loss": 0.3442,
+      "step": 596
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.336832835156173,
+      "learning_rate": 9.171966984569944e-06,
+      "loss": 0.3984,
+      "step": 597
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.923069020600507,
+      "learning_rate": 9.168815990896654e-06,
+      "loss": 0.3662,
+      "step": 598
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.187383539442544,
+      "learning_rate": 9.165659556620365e-06,
+      "loss": 0.3867,
+      "step": 599
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.442481752021603,
+      "learning_rate": 9.162497685860449e-06,
+      "loss": 0.436,
+      "step": 600
+    },
+    {
+      "epoch": 0.85,
+      "eval_avg_AUC": 0.7436113836582854,
+      "eval_avg_Accuracy": 0.6419098143236074,
+      "eval_avg_Accuracy-right": 0.9956306247554454,
+      "eval_avg_Accuracy-wrong": 0.025130770980213782,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.621815969426199,
+      "eval_last_AUC": 0.7523205228531411,
+      "eval_last_Accuracy": 0.7048657161803713,
+      "eval_last_Accuracy-right": 0.87022303378114,
+      "eval_last_Accuracy-wrong": 0.41653400045485556,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6237919767353252,
+      "eval_max_AUC": 0.5296710841506856,
+      "eval_max_Accuracy": 0.6358173076923077,
+      "eval_max_Accuracy-right": 0.999478283552889,
+      "eval_max_Accuracy-wrong": 0.0017057084375710712,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5348953658957699,
+      "eval_min_AUC": 0.7482432782658677,
+      "eval_min_Accuracy": 0.7065649867374005,
+      "eval_min_Accuracy-right": 0.8671579496543629,
+      "eval_min_Accuracy-wrong": 0.4265408232886059,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6205421612944162,
+      "eval_prod_AUC": 0.7461364565563529,
+      "eval_prod_Accuracy": 0.6409980106100795,
+      "eval_prod_Accuracy-right": 0.5455197600104343,
+      "eval_prod_Accuracy-wrong": 0.8074823743461451,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6188953480320266,
+      "eval_runtime": 667.6754,
+      "eval_samples_per_second": 36.137,
+      "eval_steps_per_second": 1.129,
+      "eval_sum_AUC": 0.5338000947383437,
+      "eval_sum_Accuracy": 0.6356929708222812,
+      "eval_sum_Accuracy-right": 1.0,
+      "eval_sum_Accuracy-wrong": 0.00045485558335228563,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.5949748204276974,
+      "step": 600
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.172261139495567,
+      "learning_rate": 9.159330382743375e-06,
+      "loss": 0.3584,
+      "step": 601
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.522543865401317,
+      "learning_rate": 9.1561576514027e-06,
+      "loss": 0.4326,
+      "step": 602
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8823743051804596,
+      "learning_rate": 9.152979495979064e-06,
+      "loss": 0.4429,
+      "step": 603
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.51962214398988,
+      "learning_rate": 9.149795920620188e-06,
+      "loss": 0.3206,
+      "step": 604
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.793045092390943,
+      "learning_rate": 9.146606929480866e-06,
+      "loss": 0.3567,
+      "step": 605
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.054689296680376,
+      "learning_rate": 9.143412526722958e-06,
+      "loss": 0.3354,
+      "step": 606
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5279729613010877,
+      "learning_rate": 9.140212716515391e-06,
+      "loss": 0.3662,
+      "step": 607
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.733473418136833,
+      "learning_rate": 9.137007503034141e-06,
+      "loss": 0.3298,
+      "step": 608
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.232683136870465,
+      "learning_rate": 9.133796890462245e-06,
+      "loss": 0.313,
+      "step": 609
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.143600346604078,
+      "learning_rate": 9.130580882989783e-06,
+      "loss": 0.4224,
+      "step": 610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.267773773965147,
+      "learning_rate": 9.12735948481387e-06,
+      "loss": 0.2909,
+      "step": 611
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.364441463210925,
+      "learning_rate": 9.124132700138666e-06,
+      "loss": 0.4514,
+      "step": 612
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.164562447501559,
+      "learning_rate": 9.120900533175352e-06,
+      "loss": 0.3838,
+      "step": 613
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.0162917080032186,
+      "learning_rate": 9.117662988142138e-06,
+      "loss": 0.2854,
+      "step": 614
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.2587929849686805,
+      "learning_rate": 9.114420069264252e-06,
+      "loss": 0.3093,
+      "step": 615
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8643016173037616,
+      "learning_rate": 9.111171780773938e-06,
+      "loss": 0.3513,
+      "step": 616
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.85843173734757,
+      "learning_rate": 9.107918126910439e-06,
+      "loss": 0.3711,
+      "step": 617
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.854288270009519,
+      "learning_rate": 9.10465911192001e-06,
+      "loss": 0.3345,
+      "step": 618
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.44714032968495,
+      "learning_rate": 9.101394740055897e-06,
+      "loss": 0.2983,
+      "step": 619
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.71266646355715,
+      "learning_rate": 9.09812501557834e-06,
+      "loss": 0.2876,
+      "step": 620
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.073195480642503,
+      "learning_rate": 9.094849942754564e-06,
+      "loss": 0.3569,
+      "step": 621
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.778811389041055,
+      "learning_rate": 9.091569525858772e-06,
+      "loss": 0.3296,
+      "step": 622
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.348171969259551,
+      "learning_rate": 9.088283769172141e-06,
+      "loss": 0.3491,
+      "step": 623
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.12550620494004,
+      "learning_rate": 9.08499267698282e-06,
+      "loss": 0.437,
+      "step": 624
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.0670447789350535,
+      "learning_rate": 9.08169625358592e-06,
+      "loss": 0.3135,
+      "step": 625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.238235419746063,
+      "learning_rate": 9.078394503283509e-06,
+      "loss": 0.5005,
+      "step": 626
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.324757066041814,
+      "learning_rate": 9.075087430384608e-06,
+      "loss": 0.3921,
+      "step": 627
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.177537555475432,
+      "learning_rate": 9.071775039205183e-06,
+      "loss": 0.3472,
+      "step": 628
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.082574557870084,
+      "learning_rate": 9.06845733406814e-06,
+      "loss": 0.3882,
+      "step": 629
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.466229500976798,
+      "learning_rate": 9.065134319303324e-06,
+      "loss": 0.395,
+      "step": 630
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9100826404230618,
+      "learning_rate": 9.061805999247504e-06,
+      "loss": 0.3687,
+      "step": 631
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.537422050338189,
+      "learning_rate": 9.058472378244378e-06,
+      "loss": 0.2737,
+      "step": 632
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.621086871364713,
+      "learning_rate": 9.055133460644561e-06,
+      "loss": 0.3555,
+      "step": 633
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.058315503047276,
+      "learning_rate": 9.051789250805575e-06,
+      "loss": 0.2793,
+      "step": 634
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.827356457021118,
+      "learning_rate": 9.048439753091855e-06,
+      "loss": 0.3147,
+      "step": 635
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.284299015139685,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.3445,
+      "step": 636
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4275726820854713,
+      "learning_rate": 9.04172491153245e-06,
+      "loss": 0.2712,
+      "step": 637
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.865240435835414,
+      "learning_rate": 9.03835957645011e-06,
+      "loss": 0.2378,
+      "step": 638
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.358115905575971,
+      "learning_rate": 9.034988971019722e-06,
+      "loss": 0.3298,
+      "step": 639
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.840812403330787,
+      "learning_rate": 9.031613099640167e-06,
+      "loss": 0.3281,
+      "step": 640
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8394564424556745,
+      "learning_rate": 9.0282319667172e-06,
+      "loss": 0.3062,
+      "step": 641
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.592159060669056,
+      "learning_rate": 9.024845576663437e-06,
+      "loss": 0.332,
+      "step": 642
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.577044839013437,
+      "learning_rate": 9.021453933898364e-06,
+      "loss": 0.4165,
+      "step": 643
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.650147232333098,
+      "learning_rate": 9.018057042848315e-06,
+      "loss": 0.2729,
+      "step": 644
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.343384349986136,
+      "learning_rate": 9.014654907946476e-06,
+      "loss": 0.2847,
+      "step": 645
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.548303921863648,
+      "learning_rate": 9.011247533632876e-06,
+      "loss": 0.2886,
+      "step": 646
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.042880454641808,
+      "learning_rate": 9.007834924354384e-06,
+      "loss": 0.3452,
+      "step": 647
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.298447039162743,
+      "learning_rate": 9.004417084564696e-06,
+      "loss": 0.3049,
+      "step": 648
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.074128589369922,
+      "learning_rate": 9.00099401872434e-06,
+      "loss": 0.2808,
+      "step": 649
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.527003504341122,
+      "learning_rate": 8.997565731300661e-06,
+      "loss": 0.2959,
+      "step": 650
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.835237566036188,
+      "learning_rate": 8.99413222676782e-06,
+      "loss": 0.4517,
+      "step": 651
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.489053972476339,
+      "learning_rate": 8.990693509606782e-06,
+      "loss": 0.3896,
+      "step": 652
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.426891658441978,
+      "learning_rate": 8.987249584305326e-06,
+      "loss": 0.3374,
+      "step": 653
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.547246221838683,
+      "learning_rate": 8.983800455358014e-06,
+      "loss": 0.3042,
+      "step": 654
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.2470148601061615,
+      "learning_rate": 8.98034612726621e-06,
+      "loss": 0.3748,
+      "step": 655
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.766538202088519,
+      "learning_rate": 8.976886604538055e-06,
+      "loss": 0.3872,
+      "step": 656
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.076227843742269,
+      "learning_rate": 8.973421891688479e-06,
+      "loss": 0.333,
+      "step": 657
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5443913474604822,
+      "learning_rate": 8.969951993239177e-06,
+      "loss": 0.3325,
+      "step": 658
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1158320149737935,
+      "learning_rate": 8.966476913718615e-06,
+      "loss": 0.3408,
+      "step": 659
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.985749241931523,
+      "learning_rate": 8.96299665766202e-06,
+      "loss": 0.2776,
+      "step": 660
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7502657646500874,
+      "learning_rate": 8.959511229611377e-06,
+      "loss": 0.2512,
+      "step": 661
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.808440155553063,
+      "learning_rate": 8.956020634115415e-06,
+      "loss": 0.2954,
+      "step": 662
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.107973569750316,
+      "learning_rate": 8.952524875729614e-06,
+      "loss": 0.3914,
+      "step": 663
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.239339161556028,
+      "learning_rate": 8.949023959016188e-06,
+      "loss": 0.3765,
+      "step": 664
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.833793975738804,
+      "learning_rate": 8.945517888544081e-06,
+      "loss": 0.3765,
+      "step": 665
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.721120184137981,
+      "learning_rate": 8.942006668888972e-06,
+      "loss": 0.3472,
+      "step": 666
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.003088043342945,
+      "learning_rate": 8.938490304633247e-06,
+      "loss": 0.458,
+      "step": 667
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.360538500943193,
+      "learning_rate": 8.934968800366017e-06,
+      "loss": 0.3467,
+      "step": 668
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.082351849293712,
+      "learning_rate": 8.931442160683094e-06,
+      "loss": 0.4028,
+      "step": 669
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.0090657255737545,
+      "learning_rate": 8.927910390187e-06,
+      "loss": 0.353,
+      "step": 670
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.097740638905433,
+      "learning_rate": 8.924373493486941e-06,
+      "loss": 0.3333,
+      "step": 671
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.727015324005116,
+      "learning_rate": 8.920831475198825e-06,
+      "loss": 0.3511,
+      "step": 672
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.583472918475056,
+      "learning_rate": 8.917284339945238e-06,
+      "loss": 0.293,
+      "step": 673
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.28807645490256,
+      "learning_rate": 8.913732092355445e-06,
+      "loss": 0.4707,
+      "step": 674
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.307365603567834,
+      "learning_rate": 8.910174737065384e-06,
+      "loss": 0.3984,
+      "step": 675
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.442401342536438,
+      "learning_rate": 8.906612278717657e-06,
+      "loss": 0.3042,
+      "step": 676
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9852383813710563,
+      "learning_rate": 8.903044721961526e-06,
+      "loss": 0.2466,
+      "step": 677
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6928998785741083,
+      "learning_rate": 8.89947207145291e-06,
+      "loss": 0.2959,
+      "step": 678
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.420918261951631,
+      "learning_rate": 8.89589433185437e-06,
+      "loss": 0.3032,
+      "step": 679
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.913052354587028,
+      "learning_rate": 8.892311507835118e-06,
+      "loss": 0.3403,
+      "step": 680
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5583736508142243,
+      "learning_rate": 8.88872360407099e-06,
+      "loss": 0.3179,
+      "step": 681
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.8623466948256775,
+      "learning_rate": 8.885130625244459e-06,
+      "loss": 0.4019,
+      "step": 682
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.278954889391585,
+      "learning_rate": 8.881532576044621e-06,
+      "loss": 0.3547,
+      "step": 683
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.277052244006082,
+      "learning_rate": 8.877929461167182e-06,
+      "loss": 0.3696,
+      "step": 684
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.556083361456363,
+      "learning_rate": 8.874321285314471e-06,
+      "loss": 0.3721,
+      "step": 685
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.300063364940838,
+      "learning_rate": 8.870708053195414e-06,
+      "loss": 0.3389,
+      "step": 686
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.395280263370811,
+      "learning_rate": 8.867089769525535e-06,
+      "loss": 0.3579,
+      "step": 687
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.318703679999736,
+      "learning_rate": 8.863466439026952e-06,
+      "loss": 0.3452,
+      "step": 688
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7711757380928987,
+      "learning_rate": 8.859838066428374e-06,
+      "loss": 0.3237,
+      "step": 689
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.609606395696577,
+      "learning_rate": 8.856204656465085e-06,
+      "loss": 0.2488,
+      "step": 690
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.55618101785313,
+      "learning_rate": 8.852566213878947e-06,
+      "loss": 0.3145,
+      "step": 691
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.907450878240477,
+      "learning_rate": 8.848922743418384e-06,
+      "loss": 0.3525,
+      "step": 692
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.195439965135572,
+      "learning_rate": 8.845274249838389e-06,
+      "loss": 0.3435,
+      "step": 693
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.675714529711746,
+      "learning_rate": 8.841620737900504e-06,
+      "loss": 0.4863,
+      "step": 694
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.833053287933571,
+      "learning_rate": 8.837962212372825e-06,
+      "loss": 0.3613,
+      "step": 695
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.515562784298242,
+      "learning_rate": 8.834298678029988e-06,
+      "loss": 0.4482,
+      "step": 696
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.500786625444841,
+      "learning_rate": 8.830630139653172e-06,
+      "loss": 0.373,
+      "step": 697
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.673530467281575,
+      "learning_rate": 8.826956602030077e-06,
+      "loss": 0.3525,
+      "step": 698
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.643645419453327,
+      "learning_rate": 8.823278069954932e-06,
+      "loss": 0.4224,
+      "step": 699
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.730147362078206,
+      "learning_rate": 8.81959454822849e-06,
+      "loss": 0.3506,
+      "step": 700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9011495008846175,
+      "learning_rate": 8.815906041658001e-06,
+      "loss": 0.3809,
+      "step": 701
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.201506009760648,
+      "learning_rate": 8.81221255505724e-06,
+      "loss": 0.4053,
+      "step": 702
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.097798876078893,
+      "learning_rate": 8.808514093246466e-06,
+      "loss": 0.3398,
+      "step": 703
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.789769896106429,
+      "learning_rate": 8.804810661052433e-06,
+      "loss": 0.3794,
+      "step": 704
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.767978259598211,
+      "learning_rate": 8.801102263308393e-06,
+      "loss": 0.3315,
+      "step": 705
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.678271689161004,
+      "learning_rate": 8.797388904854064e-06,
+      "loss": 0.2576,
+      "step": 706
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.725812591663642,
+      "learning_rate": 8.793670590535646e-06,
+      "loss": 0.4297,
+      "step": 707
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.3381416807655673,
+      "learning_rate": 8.789947325205805e-06,
+      "loss": 0.353,
+      "step": 708
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.849364929290131,
+      "learning_rate": 8.786219113723671e-06,
+      "loss": 0.2542,
+      "step": 709
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.074601042809408,
+      "learning_rate": 8.782485960954821e-06,
+      "loss": 0.3613,
+      "step": 710
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.316488510824646,
+      "learning_rate": 8.778747871771293e-06,
+      "loss": 0.3108,
+      "step": 711
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.724712207051525,
+      "learning_rate": 8.775004851051555e-06,
+      "loss": 0.2571,
+      "step": 712
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.796991976725225,
+      "learning_rate": 8.77125690368052e-06,
+      "loss": 0.2368,
+      "step": 713
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.8673385881658735,
+      "learning_rate": 8.767504034549525e-06,
+      "loss": 0.2944,
+      "step": 714
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.1775426816463463,
+      "learning_rate": 8.763746248556338e-06,
+      "loss": 0.3184,
+      "step": 715
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.7484303239335413,
+      "learning_rate": 8.759983550605132e-06,
+      "loss": 0.2783,
+      "step": 716
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.011370520084286,
+      "learning_rate": 8.756215945606498e-06,
+      "loss": 0.3552,
+      "step": 717
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.492619246259563,
+      "learning_rate": 8.752443438477434e-06,
+      "loss": 0.2603,
+      "step": 718
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.929048880819835,
+      "learning_rate": 8.748666034141327e-06,
+      "loss": 0.2231,
+      "step": 719
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.436827863054401,
+      "learning_rate": 8.744883737527962e-06,
+      "loss": 0.2292,
+      "step": 720
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.203636350375588,
+      "learning_rate": 8.741096553573506e-06,
+      "loss": 0.2996,
+      "step": 721
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.3944698570205265,
+      "learning_rate": 8.737304487220508e-06,
+      "loss": 0.3496,
+      "step": 722
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.439043968021188,
+      "learning_rate": 8.733507543417881e-06,
+      "loss": 0.3103,
+      "step": 723
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.860763756627871,
+      "learning_rate": 8.729705727120911e-06,
+      "loss": 0.2341,
+      "step": 724
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.796023573523343,
+      "learning_rate": 8.72589904329124e-06,
+      "loss": 0.3022,
+      "step": 725
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.78756042674705,
+      "learning_rate": 8.72208749689686e-06,
+      "loss": 0.271,
+      "step": 726
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.533211452516399,
+      "learning_rate": 8.718271092912116e-06,
+      "loss": 0.2812,
+      "step": 727
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.0666651830124225,
+      "learning_rate": 8.714449836317684e-06,
+      "loss": 0.3088,
+      "step": 728
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.084707455818183,
+      "learning_rate": 8.71062373210058e-06,
+      "loss": 0.3569,
+      "step": 729
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.537692569985042,
+      "learning_rate": 8.706792785254141e-06,
+      "loss": 0.3884,
+      "step": 730
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.210607020154836,
+      "learning_rate": 8.702957000778029e-06,
+      "loss": 0.2876,
+      "step": 731
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.522862837670095,
+      "learning_rate": 8.699116383678219e-06,
+      "loss": 0.2559,
+      "step": 732
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.449639797542584,
+      "learning_rate": 8.695270938966984e-06,
+      "loss": 0.2141,
+      "step": 733
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 51.91385486387932,
+      "learning_rate": 8.69142067166291e-06,
+      "loss": 0.2866,
+      "step": 734
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.330323109839878,
+      "learning_rate": 8.68756558679087e-06,
+      "loss": 0.2832,
+      "step": 735
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 46.01839169809924,
+      "learning_rate": 8.683705689382025e-06,
+      "loss": 0.3357,
+      "step": 736
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 34.797262290100825,
+      "learning_rate": 8.679840984473817e-06,
+      "loss": 0.3522,
+      "step": 737
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.610912460827764,
+      "learning_rate": 8.675971477109963e-06,
+      "loss": 0.262,
+      "step": 738
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.074209429327233,
+      "learning_rate": 8.672097172340449e-06,
+      "loss": 0.2549,
+      "step": 739
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 21.672675238163734,
+      "learning_rate": 8.668218075221513e-06,
+      "loss": 0.3013,
+      "step": 740
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.755256434137868,
+      "learning_rate": 8.66433419081566e-06,
+      "loss": 0.2163,
+      "step": 741
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.289910398692217,
+      "learning_rate": 8.660445524191636e-06,
+      "loss": 0.2666,
+      "step": 742
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 33.31973521661891,
+      "learning_rate": 8.656552080424428e-06,
+      "loss": 0.5522,
+      "step": 743
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.760120362902158,
+      "learning_rate": 8.652653864595258e-06,
+      "loss": 0.2739,
+      "step": 744
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.733281705803904,
+      "learning_rate": 8.648750881791577e-06,
+      "loss": 0.2961,
+      "step": 745
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.10500694305235,
+      "learning_rate": 8.644843137107058e-06,
+      "loss": 0.3013,
+      "step": 746
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.950833577819047,
+      "learning_rate": 8.640930635641587e-06,
+      "loss": 0.2676,
+      "step": 747
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.352590440023834,
+      "learning_rate": 8.637013382501256e-06,
+      "loss": 0.3149,
+      "step": 748
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.947018121820923,
+      "learning_rate": 8.633091382798362e-06,
+      "loss": 0.3398,
+      "step": 749
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.095441628747481,
+      "learning_rate": 8.629164641651397e-06,
+      "loss": 0.2744,
+      "step": 750
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.008649193478815,
+      "learning_rate": 8.625233164185035e-06,
+      "loss": 0.3279,
+      "step": 751
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 19.912305580166255,
+      "learning_rate": 8.621296955530138e-06,
+      "loss": 0.4766,
+      "step": 752
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.27852950890214,
+      "learning_rate": 8.617356020823739e-06,
+      "loss": 0.2263,
+      "step": 753
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.008033098285212,
+      "learning_rate": 8.61341036520904e-06,
+      "loss": 0.3093,
+      "step": 754
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.994152245312076,
+      "learning_rate": 8.609459993835404e-06,
+      "loss": 0.24,
+      "step": 755
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.960293571513324,
+      "learning_rate": 8.605504911858347e-06,
+      "loss": 0.2644,
+      "step": 756
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 13.98472524473993,
+      "learning_rate": 8.601545124439535e-06,
+      "loss": 0.2869,
+      "step": 757
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.714312963643217,
+      "learning_rate": 8.597580636746773e-06,
+      "loss": 0.2456,
+      "step": 758
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.989063248003022,
+      "learning_rate": 8.593611453954e-06,
+      "loss": 0.3384,
+      "step": 759
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.039660662226755,
+      "learning_rate": 8.589637581241287e-06,
+      "loss": 0.238,
+      "step": 760
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.579837428740737,
+      "learning_rate": 8.585659023794818e-06,
+      "loss": 0.302,
+      "step": 761
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.537627988342262,
+      "learning_rate": 8.581675786806898e-06,
+      "loss": 0.2549,
+      "step": 762
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 14.250559209049381,
+      "learning_rate": 8.577687875475933e-06,
+      "loss": 0.2576,
+      "step": 763
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 15.76921131661366,
+      "learning_rate": 8.573695295006434e-06,
+      "loss": 0.3428,
+      "step": 764
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.97604909831586,
+      "learning_rate": 8.569698050609007e-06,
+      "loss": 0.2568,
+      "step": 765
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.292274777831013,
+      "learning_rate": 8.565696147500338e-06,
+      "loss": 0.2688,
+      "step": 766
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.511306596960287,
+      "learning_rate": 8.561689590903197e-06,
+      "loss": 0.2754,
+      "step": 767
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.698104382525688,
+      "learning_rate": 8.557678386046429e-06,
+      "loss": 0.271,
+      "step": 768
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.600255012445068,
+      "learning_rate": 8.55366253816494e-06,
+      "loss": 0.4121,
+      "step": 769
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.7617732937276,
+      "learning_rate": 8.549642052499702e-06,
+      "loss": 0.2446,
+      "step": 770
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.277994087516516,
+      "learning_rate": 8.545616934297733e-06,
+      "loss": 0.2944,
+      "step": 771
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.702731539691897,
+      "learning_rate": 8.541587188812103e-06,
+      "loss": 0.2622,
+      "step": 772
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.980597493860651,
+      "learning_rate": 8.537552821301919e-06,
+      "loss": 0.2517,
+      "step": 773
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.65012932032952,
+      "learning_rate": 8.533513837032317e-06,
+      "loss": 0.3298,
+      "step": 774
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.24364821790684,
+      "learning_rate": 8.529470241274458e-06,
+      "loss": 0.325,
+      "step": 775
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.042174546756993,
+      "learning_rate": 8.525422039305529e-06,
+      "loss": 0.2869,
+      "step": 776
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.66757263356849,
+      "learning_rate": 8.521369236408718e-06,
+      "loss": 0.3088,
+      "step": 777
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.797448574169143,
+      "learning_rate": 8.517311837873225e-06,
+      "loss": 0.3284,
+      "step": 778
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.852399332342742,
+      "learning_rate": 8.513249848994248e-06,
+      "loss": 0.3196,
+      "step": 779
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.78125307546963,
+      "learning_rate": 8.509183275072967e-06,
+      "loss": 0.3022,
+      "step": 780
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.346655540118093,
+      "learning_rate": 8.505112121416554e-06,
+      "loss": 0.2471,
+      "step": 781
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.035793115539333,
+      "learning_rate": 8.50103639333816e-06,
+      "loss": 0.2561,
+      "step": 782
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.445091918889698,
+      "learning_rate": 8.496956096156896e-06,
+      "loss": 0.3333,
+      "step": 783
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.965391747286695,
+      "learning_rate": 8.492871235197844e-06,
+      "loss": 0.2386,
+      "step": 784
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 19.532364156971322,
+      "learning_rate": 8.48878181579204e-06,
+      "loss": 0.3867,
+      "step": 785
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.159290258485411,
+      "learning_rate": 8.48468784327647e-06,
+      "loss": 0.3389,
+      "step": 786
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.569672212074234,
+      "learning_rate": 8.480589322994056e-06,
+      "loss": 0.3657,
+      "step": 787
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.65729714894695,
+      "learning_rate": 8.476486260293664e-06,
+      "loss": 0.3452,
+      "step": 788
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.240084521699,
+      "learning_rate": 8.472378660530082e-06,
+      "loss": 0.3887,
+      "step": 789
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.694396319428277,
+      "learning_rate": 8.468266529064025e-06,
+      "loss": 0.2993,
+      "step": 790
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.271410133674335,
+      "learning_rate": 8.464149871262118e-06,
+      "loss": 0.3127,
+      "step": 791
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.952604117455994,
+      "learning_rate": 8.460028692496889e-06,
+      "loss": 0.2969,
+      "step": 792
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 14.046350177171979,
+      "learning_rate": 8.455902998146776e-06,
+      "loss": 0.4136,
+      "step": 793
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.336748919812557,
+      "learning_rate": 8.451772793596105e-06,
+      "loss": 0.2449,
+      "step": 794
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.287304973294653,
+      "learning_rate": 8.447638084235083e-06,
+      "loss": 0.2954,
+      "step": 795
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.197897186616546,
+      "learning_rate": 8.44349887545981e-06,
+      "loss": 0.3213,
+      "step": 796
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.061614735896679,
+      "learning_rate": 8.439355172672241e-06,
+      "loss": 0.3118,
+      "step": 797
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.403231472447587,
+      "learning_rate": 8.435206981280209e-06,
+      "loss": 0.293,
+      "step": 798
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.753974587614309,
+      "learning_rate": 8.431054306697402e-06,
+      "loss": 0.3169,
+      "step": 799
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.062372064743458,
+      "learning_rate": 8.426897154343353e-06,
+      "loss": 0.2227,
+      "step": 800
+    },
+    {
+      "epoch": 1.13,
+      "eval_avg_AUC": 0.7117075537855455,
+      "eval_avg_Accuracy": 0.6416611405835544,
+      "eval_avg_Accuracy-right": 0.9701969479587844,
+      "eval_avg_Accuracy-wrong": 0.06879690698203321,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6269866405639123,
+      "eval_last_AUC": 0.7416692112752571,
+      "eval_last_Accuracy": 0.7056531830238727,
+      "eval_last_Accuracy-right": 0.83487674448937,
+      "eval_last_Accuracy-wrong": 0.4803274960200136,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6249021890282285,
+      "eval_max_AUC": 0.5122088671387725,
+      "eval_max_Accuracy": 0.6365633289124668,
+      "eval_max_Accuracy-right": 0.9949784791965567,
+      "eval_max_Accuracy-wrong": 0.011598817375483284,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5054396695832021,
+      "eval_min_AUC": 0.7390329606853672,
+      "eval_min_Accuracy": 0.7053630636604774,
+      "eval_min_Accuracy-right": 0.823464197208817,
+      "eval_min_Accuracy-wrong": 0.4994314305208096,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6230229558248375,
+      "eval_prod_AUC": 0.7316445868106891,
+      "eval_prod_Accuracy": 0.6567887931034483,
+      "eval_prod_Accuracy-right": 0.60486500586931,
+      "eval_prod_Accuracy-wrong": 0.7473277234478053,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.620981620630729,
+      "eval_runtime": 668.284,
+      "eval_samples_per_second": 36.104,
+      "eval_steps_per_second": 1.128,
+      "eval_sum_AUC": 0.5309207022740021,
+      "eval_sum_Accuracy": 0.6363975464190982,
+      "eval_sum_Accuracy-right": 0.999478283552889,
+      "eval_sum_Accuracy-wrong": 0.003297702979304071,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.5915878456085056,
+      "step": 800
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.661523639419524,
+      "learning_rate": 8.422735529643445e-06,
+      "loss": 0.3647,
+      "step": 801
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.429991237297951,
+      "learning_rate": 8.418569438028897e-06,
+      "loss": 0.2429,
+      "step": 802
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.8725219544719,
+      "learning_rate": 8.414398884936752e-06,
+      "loss": 0.3376,
+      "step": 803
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.948876813509901,
+      "learning_rate": 8.410223875809886e-06,
+      "loss": 0.3892,
+      "step": 804
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.609326557600053,
+      "learning_rate": 8.406044416096983e-06,
+      "loss": 0.3657,
+      "step": 805
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.281320826635043,
+      "learning_rate": 8.401860511252535e-06,
+      "loss": 0.2268,
+      "step": 806
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.571204534446798,
+      "learning_rate": 8.397672166736836e-06,
+      "loss": 0.2378,
+      "step": 807
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.467254900983802,
+      "learning_rate": 8.393479388015979e-06,
+      "loss": 0.3828,
+      "step": 808
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.781243967177545,
+      "learning_rate": 8.389282180561837e-06,
+      "loss": 0.2749,
+      "step": 809
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.500417110792991,
+      "learning_rate": 8.385080549852067e-06,
+      "loss": 0.2634,
+      "step": 810
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.238228964385284,
+      "learning_rate": 8.380874501370098e-06,
+      "loss": 0.2665,
+      "step": 811
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.002403278189231,
+      "learning_rate": 8.376664040605122e-06,
+      "loss": 0.2832,
+      "step": 812
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.329085308031678,
+      "learning_rate": 8.372449173052096e-06,
+      "loss": 0.3716,
+      "step": 813
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.265536582099523,
+      "learning_rate": 8.368229904211718e-06,
+      "loss": 0.2932,
+      "step": 814
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.367706332332974,
+      "learning_rate": 8.364006239590438e-06,
+      "loss": 0.2488,
+      "step": 815
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.165537179708043,
+      "learning_rate": 8.35977818470044e-06,
+      "loss": 0.3418,
+      "step": 816
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.145313409957077,
+      "learning_rate": 8.35554574505964e-06,
+      "loss": 0.2786,
+      "step": 817
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.854906745240785,
+      "learning_rate": 8.351308926191672e-06,
+      "loss": 0.238,
+      "step": 818
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.761962689988771,
+      "learning_rate": 8.34706773362589e-06,
+      "loss": 0.3274,
+      "step": 819
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.913392146053789,
+      "learning_rate": 8.342822172897349e-06,
+      "loss": 0.3086,
+      "step": 820
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.451164847598736,
+      "learning_rate": 8.338572249546813e-06,
+      "loss": 0.3884,
+      "step": 821
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.503978528732752,
+      "learning_rate": 8.334317969120738e-06,
+      "loss": 0.3237,
+      "step": 822
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.727330679364709,
+      "learning_rate": 8.33005933717126e-06,
+      "loss": 0.2498,
+      "step": 823
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.1766201878733,
+      "learning_rate": 8.325796359256199e-06,
+      "loss": 0.3367,
+      "step": 824
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.597955877752988,
+      "learning_rate": 8.321529040939046e-06,
+      "loss": 0.2981,
+      "step": 825
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.234333809855164,
+      "learning_rate": 8.31725738778896e-06,
+      "loss": 0.3198,
+      "step": 826
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.476768525083362,
+      "learning_rate": 8.312981405380749e-06,
+      "loss": 0.3647,
+      "step": 827
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.619446295442009,
+      "learning_rate": 8.308701099294879e-06,
+      "loss": 0.3091,
+      "step": 828
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.762643950697453,
+      "learning_rate": 8.304416475117457e-06,
+      "loss": 0.3298,
+      "step": 829
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.06377121115096,
+      "learning_rate": 8.300127538440218e-06,
+      "loss": 0.4448,
+      "step": 830
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.75303077888266,
+      "learning_rate": 8.295834294860535e-06,
+      "loss": 0.3516,
+      "step": 831
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.7505559535106,
+      "learning_rate": 8.2915367499814e-06,
+      "loss": 0.3198,
+      "step": 832
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.796224077716893,
+      "learning_rate": 8.287234909411412e-06,
+      "loss": 0.3591,
+      "step": 833
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.1498678384093255,
+      "learning_rate": 8.282928778764783e-06,
+      "loss": 0.2659,
+      "step": 834
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 18.960808659579417,
+      "learning_rate": 8.278618363661322e-06,
+      "loss": 0.5015,
+      "step": 835
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.374189247807947,
+      "learning_rate": 8.274303669726427e-06,
+      "loss": 0.3074,
+      "step": 836
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.67522962265373,
+      "learning_rate": 8.269984702591082e-06,
+      "loss": 0.2605,
+      "step": 837
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.480093182741655,
+      "learning_rate": 8.26566146789185e-06,
+      "loss": 0.3311,
+      "step": 838
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.1420512491682,
+      "learning_rate": 8.261333971270861e-06,
+      "loss": 0.2617,
+      "step": 839
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.064215671935924,
+      "learning_rate": 8.257002218375807e-06,
+      "loss": 0.2673,
+      "step": 840
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.321949943904182,
+      "learning_rate": 8.252666214859936e-06,
+      "loss": 0.3107,
+      "step": 841
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.89673739211508,
+      "learning_rate": 8.24832596638204e-06,
+      "loss": 0.3804,
+      "step": 842
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.511606479358201,
+      "learning_rate": 8.243981478606459e-06,
+      "loss": 0.283,
+      "step": 843
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.779036034278954,
+      "learning_rate": 8.239632757203056e-06,
+      "loss": 0.4424,
+      "step": 844
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.874437070937855,
+      "learning_rate": 8.235279807847223e-06,
+      "loss": 0.3374,
+      "step": 845
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.162804979402527,
+      "learning_rate": 8.230922636219872e-06,
+      "loss": 0.2991,
+      "step": 846
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.273020666742799,
+      "learning_rate": 8.226561248007424e-06,
+      "loss": 0.3196,
+      "step": 847
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.093760443008431,
+      "learning_rate": 8.2221956489018e-06,
+      "loss": 0.2563,
+      "step": 848
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.310939576343992,
+      "learning_rate": 8.217825844600422e-06,
+      "loss": 0.4307,
+      "step": 849
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.5236039921185185,
+      "learning_rate": 8.213451840806197e-06,
+      "loss": 0.3713,
+      "step": 850
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.507123423860703,
+      "learning_rate": 8.20907364322751e-06,
+      "loss": 0.3789,
+      "step": 851
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.439242335218645,
+      "learning_rate": 8.204691257578225e-06,
+      "loss": 0.3171,
+      "step": 852
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.561269525777554,
+      "learning_rate": 8.200304689577665e-06,
+      "loss": 0.3105,
+      "step": 853
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 23.625155940322212,
+      "learning_rate": 8.19591394495062e-06,
+      "loss": 0.4067,
+      "step": 854
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.80272141870214,
+      "learning_rate": 8.19151902942732e-06,
+      "loss": 0.3755,
+      "step": 855
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.509708308954935,
+      "learning_rate": 8.18711994874345e-06,
+      "loss": 0.3132,
+      "step": 856
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.987739857629714,
+      "learning_rate": 8.18271670864012e-06,
+      "loss": 0.3098,
+      "step": 857
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.282590757760539,
+      "learning_rate": 8.178309314863875e-06,
+      "loss": 0.2166,
+      "step": 858
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.089239445549541,
+      "learning_rate": 8.173897773166678e-06,
+      "loss": 0.2856,
+      "step": 859
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.240812953115288,
+      "learning_rate": 8.169482089305912e-06,
+      "loss": 0.2617,
+      "step": 860
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.745981396884222,
+      "learning_rate": 8.165062269044353e-06,
+      "loss": 0.3335,
+      "step": 861
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.212950235018874,
+      "learning_rate": 8.160638318150185e-06,
+      "loss": 0.2468,
+      "step": 862
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.5242398739806955,
+      "learning_rate": 8.156210242396981e-06,
+      "loss": 0.2363,
+      "step": 863
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.59833371659614,
+      "learning_rate": 8.151778047563697e-06,
+      "loss": 0.3936,
+      "step": 864
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.250602948785272,
+      "learning_rate": 8.147341739434662e-06,
+      "loss": 0.3469,
+      "step": 865
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.811068139364268,
+      "learning_rate": 8.142901323799578e-06,
+      "loss": 0.3081,
+      "step": 866
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.239231487338333,
+      "learning_rate": 8.138456806453503e-06,
+      "loss": 0.2705,
+      "step": 867
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.628630570961177,
+      "learning_rate": 8.134008193196852e-06,
+      "loss": 0.3706,
+      "step": 868
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.303082556682579,
+      "learning_rate": 8.12955548983538e-06,
+      "loss": 0.4199,
+      "step": 869
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.130316148054697,
+      "learning_rate": 8.125098702180187e-06,
+      "loss": 0.3188,
+      "step": 870
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.190131381436751,
+      "learning_rate": 8.120637836047698e-06,
+      "loss": 0.3323,
+      "step": 871
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.092039497857519,
+      "learning_rate": 8.11617289725966e-06,
+      "loss": 0.3044,
+      "step": 872
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.281337957074198,
+      "learning_rate": 8.111703891643144e-06,
+      "loss": 0.292,
+      "step": 873
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.906560129664632,
+      "learning_rate": 8.107230825030514e-06,
+      "loss": 0.3286,
+      "step": 874
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.426132616690519,
+      "learning_rate": 8.102753703259448e-06,
+      "loss": 0.3271,
+      "step": 875
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.683076450148496,
+      "learning_rate": 8.098272532172906e-06,
+      "loss": 0.2505,
+      "step": 876
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.03897356082831,
+      "learning_rate": 8.093787317619139e-06,
+      "loss": 0.2659,
+      "step": 877
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.134129464212137,
+      "learning_rate": 8.089298065451673e-06,
+      "loss": 0.4121,
+      "step": 878
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.827293617778553,
+      "learning_rate": 8.084804781529301e-06,
+      "loss": 0.4253,
+      "step": 879
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.123516939747833,
+      "learning_rate": 8.080307471716083e-06,
+      "loss": 0.2881,
+      "step": 880
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.442314207975982,
+      "learning_rate": 8.075806141881327e-06,
+      "loss": 0.3496,
+      "step": 881
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.19063522301289,
+      "learning_rate": 8.07130079789959e-06,
+      "loss": 0.4214,
+      "step": 882
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.970878068556775,
+      "learning_rate": 8.06679144565067e-06,
+      "loss": 0.3784,
+      "step": 883
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.4970315750303,
+      "learning_rate": 8.062278091019593e-06,
+      "loss": 0.3779,
+      "step": 884
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.546416417066848,
+      "learning_rate": 8.057760739896612e-06,
+      "loss": 0.334,
+      "step": 885
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.46763587557673,
+      "learning_rate": 8.053239398177191e-06,
+      "loss": 0.2747,
+      "step": 886
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.4927181314315945,
+      "learning_rate": 8.048714071762005e-06,
+      "loss": 0.302,
+      "step": 887
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.273491918784087,
+      "learning_rate": 8.044184766556928e-06,
+      "loss": 0.3975,
+      "step": 888
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 17.579358962762765,
+      "learning_rate": 8.039651488473028e-06,
+      "loss": 0.4419,
+      "step": 889
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.363635262089246,
+      "learning_rate": 8.035114243426556e-06,
+      "loss": 0.2155,
+      "step": 890
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.379230922029222,
+      "learning_rate": 8.030573037338942e-06,
+      "loss": 0.3132,
+      "step": 891
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.442422020498354,
+      "learning_rate": 8.026027876136785e-06,
+      "loss": 0.3208,
+      "step": 892
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.642851555248326,
+      "learning_rate": 8.021478765751844e-06,
+      "loss": 0.4111,
+      "step": 893
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.381993218901181,
+      "learning_rate": 8.016925712121037e-06,
+      "loss": 0.2776,
+      "step": 894
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 15.080166686559245,
+      "learning_rate": 8.01236872118642e-06,
+      "loss": 0.3416,
+      "step": 895
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.426211088518768,
+      "learning_rate": 8.007807798895195e-06,
+      "loss": 0.3247,
+      "step": 896
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.235702452213214,
+      "learning_rate": 8.00324295119969e-06,
+      "loss": 0.3359,
+      "step": 897
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.470421809960511,
+      "learning_rate": 7.998674184057358e-06,
+      "loss": 0.2434,
+      "step": 898
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.913984620791439,
+      "learning_rate": 7.994101503430769e-06,
+      "loss": 0.2483,
+      "step": 899
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.45443769545427,
+      "learning_rate": 7.989524915287595e-06,
+      "loss": 0.2532,
+      "step": 900
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.571290163511465,
+      "learning_rate": 7.984944425600614e-06,
+      "loss": 0.3401,
+      "step": 901
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.396393598332807,
+      "learning_rate": 7.98036004034769e-06,
+      "loss": 0.3921,
+      "step": 902
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 6.14831116562311,
+      "learning_rate": 7.975771765511776e-06,
+      "loss": 0.2463,
+      "step": 903
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.027483868708153,
+      "learning_rate": 7.971179607080894e-06,
+      "loss": 0.2048,
+      "step": 904
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.401591267688044,
+      "learning_rate": 7.966583571048147e-06,
+      "loss": 0.3057,
+      "step": 905
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.001967668817904,
+      "learning_rate": 7.961983663411684e-06,
+      "loss": 0.2905,
+      "step": 906
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.249022365603325,
+      "learning_rate": 7.957379890174716e-06,
+      "loss": 0.3418,
+      "step": 907
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.55536597533361,
+      "learning_rate": 7.952772257345495e-06,
+      "loss": 0.3193,
+      "step": 908
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.09393203811233,
+      "learning_rate": 7.948160770937315e-06,
+      "loss": 0.312,
+      "step": 909
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.606335082058054,
+      "learning_rate": 7.943545436968491e-06,
+      "loss": 0.2786,
+      "step": 910
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.136931754498404,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.2959,
+      "step": 911
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.662629996721304,
+      "learning_rate": 7.934303250447295e-06,
+      "loss": 0.3098,
+      "step": 912
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.211033509161238,
+      "learning_rate": 7.929676409956637e-06,
+      "loss": 0.1912,
+      "step": 913
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.291770543111193,
+      "learning_rate": 7.92504574602875e-06,
+      "loss": 0.3118,
+      "step": 914
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 14.872505923036696,
+      "learning_rate": 7.92041126470698e-06,
+      "loss": 0.3589,
+      "step": 915
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.756737250741011,
+      "learning_rate": 7.91577297203966e-06,
+      "loss": 0.2595,
+      "step": 916
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 17.272422983632573,
+      "learning_rate": 7.911130874080092e-06,
+      "loss": 0.429,
+      "step": 917
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.854303304419838,
+      "learning_rate": 7.906484976886543e-06,
+      "loss": 0.4097,
+      "step": 918
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.34177571498073,
+      "learning_rate": 7.901835286522246e-06,
+      "loss": 0.302,
+      "step": 919
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.297460899955475,
+      "learning_rate": 7.897181809055377e-06,
+      "loss": 0.2905,
+      "step": 920
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.737281960867966,
+      "learning_rate": 7.892524550559056e-06,
+      "loss": 0.3247,
+      "step": 921
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.594566000055087,
+      "learning_rate": 7.887863517111337e-06,
+      "loss": 0.3306,
+      "step": 922
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.396491161905903,
+      "learning_rate": 7.883198714795208e-06,
+      "loss": 0.27,
+      "step": 923
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.1721923496353135,
+      "learning_rate": 7.878530149698565e-06,
+      "loss": 0.2981,
+      "step": 924
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.027872641361709,
+      "learning_rate": 7.873857827914221e-06,
+      "loss": 0.3876,
+      "step": 925
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.211619271974662,
+      "learning_rate": 7.869181755539888e-06,
+      "loss": 0.3347,
+      "step": 926
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.986471879570759,
+      "learning_rate": 7.864501938678177e-06,
+      "loss": 0.2903,
+      "step": 927
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 14.749235778844179,
+      "learning_rate": 7.859818383436582e-06,
+      "loss": 0.4204,
+      "step": 928
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 5.41707750287694,
+      "learning_rate": 7.85513109592748e-06,
+      "loss": 0.3115,
+      "step": 929
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.205528075591522,
+      "learning_rate": 7.850440082268112e-06,
+      "loss": 0.2424,
+      "step": 930
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.143201090834129,
+      "learning_rate": 7.845745348580592e-06,
+      "loss": 0.2817,
+      "step": 931
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.132438803998022,
+      "learning_rate": 7.841046900991878e-06,
+      "loss": 0.2988,
+      "step": 932
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.943916756454257,
+      "learning_rate": 7.836344745633785e-06,
+      "loss": 0.311,
+      "step": 933
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.859127089023723,
+      "learning_rate": 7.831638888642957e-06,
+      "loss": 0.283,
+      "step": 934
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.146979622498921,
+      "learning_rate": 7.826929336160875e-06,
+      "loss": 0.2886,
+      "step": 935
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.02719290421317,
+      "learning_rate": 7.822216094333847e-06,
+      "loss": 0.3315,
+      "step": 936
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.518933442119098,
+      "learning_rate": 7.817499169312985e-06,
+      "loss": 0.3335,
+      "step": 937
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.490611773979666,
+      "learning_rate": 7.812778567254215e-06,
+      "loss": 0.3154,
+      "step": 938
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.27632584336743,
+      "learning_rate": 7.808054294318262e-06,
+      "loss": 0.2981,
+      "step": 939
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.546057381350753,
+      "learning_rate": 7.803326356670636e-06,
+      "loss": 0.3,
+      "step": 940
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.363260954229819,
+      "learning_rate": 7.798594760481639e-06,
+      "loss": 0.2764,
+      "step": 941
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.543193271049836,
+      "learning_rate": 7.793859511926338e-06,
+      "loss": 0.2961,
+      "step": 942
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.217704711494536,
+      "learning_rate": 7.789120617184573e-06,
+      "loss": 0.3604,
+      "step": 943
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.663883711674012,
+      "learning_rate": 7.78437808244094e-06,
+      "loss": 0.3188,
+      "step": 944
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.460596871137275,
+      "learning_rate": 7.77963191388479e-06,
+      "loss": 0.2979,
+      "step": 945
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.388687784822048,
+      "learning_rate": 7.774882117710203e-06,
+      "loss": 0.343,
+      "step": 946
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.57779878862767,
+      "learning_rate": 7.770128700116011e-06,
+      "loss": 0.2834,
+      "step": 947
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.543740757393631,
+      "learning_rate": 7.76537166730576e-06,
+      "loss": 0.3616,
+      "step": 948
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.558334931478361,
+      "learning_rate": 7.760611025487721e-06,
+      "loss": 0.3533,
+      "step": 949
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.707174945962374,
+      "learning_rate": 7.75584678087487e-06,
+      "loss": 0.3328,
+      "step": 950
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.498428993556021,
+      "learning_rate": 7.751078939684886e-06,
+      "loss": 0.2842,
+      "step": 951
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.06969755517606,
+      "learning_rate": 7.746307508140145e-06,
+      "loss": 0.3118,
+      "step": 952
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.892117969995864,
+      "learning_rate": 7.741532492467705e-06,
+      "loss": 0.3101,
+      "step": 953
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.636217402258827,
+      "learning_rate": 7.736753898899302e-06,
+      "loss": 0.2424,
+      "step": 954
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.238885138351293,
+      "learning_rate": 7.731971733671347e-06,
+      "loss": 0.3845,
+      "step": 955
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.991929595224461,
+      "learning_rate": 7.727186003024902e-06,
+      "loss": 0.353,
+      "step": 956
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.450683894527396,
+      "learning_rate": 7.722396713205689e-06,
+      "loss": 0.2942,
+      "step": 957
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.483821938707923,
+      "learning_rate": 7.717603870464076e-06,
+      "loss": 0.2769,
+      "step": 958
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.734896233254307,
+      "learning_rate": 7.712807481055065e-06,
+      "loss": 0.3276,
+      "step": 959
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.928876365160588,
+      "learning_rate": 7.708007551238284e-06,
+      "loss": 0.3145,
+      "step": 960
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.120801036431401,
+      "learning_rate": 7.703204087277989e-06,
+      "loss": 0.2766,
+      "step": 961
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.598304251541482,
+      "learning_rate": 7.698397095443041e-06,
+      "loss": 0.2852,
+      "step": 962
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.927286183053114,
+      "learning_rate": 7.69358658200691e-06,
+      "loss": 0.3105,
+      "step": 963
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.229194299681224,
+      "learning_rate": 7.688772553247659e-06,
+      "loss": 0.3799,
+      "step": 964
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.658795293021148,
+      "learning_rate": 7.68395501544794e-06,
+      "loss": 0.2046,
+      "step": 965
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.38051238971842,
+      "learning_rate": 7.679133974894984e-06,
+      "loss": 0.3428,
+      "step": 966
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.798317323923194,
+      "learning_rate": 7.674309437880594e-06,
+      "loss": 0.2295,
+      "step": 967
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.173194644052108,
+      "learning_rate": 7.669481410701136e-06,
+      "loss": 0.2764,
+      "step": 968
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.843699683930156,
+      "learning_rate": 7.66464989965753e-06,
+      "loss": 0.3477,
+      "step": 969
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.538468924335463,
+      "learning_rate": 7.659814911055247e-06,
+      "loss": 0.2949,
+      "step": 970
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.682500080985811,
+      "learning_rate": 7.654976451204288e-06,
+      "loss": 0.25,
+      "step": 971
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.000663871975604,
+      "learning_rate": 7.65013452641919e-06,
+      "loss": 0.2988,
+      "step": 972
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.490521772553379,
+      "learning_rate": 7.645289143019014e-06,
+      "loss": 0.3643,
+      "step": 973
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.615152584659654,
+      "learning_rate": 7.64044030732733e-06,
+      "loss": 0.377,
+      "step": 974
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.73775258208571,
+      "learning_rate": 7.635588025672214e-06,
+      "loss": 0.261,
+      "step": 975
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.510269063494051,
+      "learning_rate": 7.630732304386244e-06,
+      "loss": 0.2727,
+      "step": 976
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.438783674433741,
+      "learning_rate": 7.6258731498064796e-06,
+      "loss": 0.2429,
+      "step": 977
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.088912935264658,
+      "learning_rate": 7.621010568274467e-06,
+      "loss": 0.3047,
+      "step": 978
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.159121327811649,
+      "learning_rate": 7.616144566136223e-06,
+      "loss": 0.2996,
+      "step": 979
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.926961144567481,
+      "learning_rate": 7.61127514974223e-06,
+      "loss": 0.2253,
+      "step": 980
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.540556105519128,
+      "learning_rate": 7.606402325447421e-06,
+      "loss": 0.3044,
+      "step": 981
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.361049221910463,
+      "learning_rate": 7.6015260996111824e-06,
+      "loss": 0.3057,
+      "step": 982
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.130017310638741,
+      "learning_rate": 7.596646478597336e-06,
+      "loss": 0.2578,
+      "step": 983
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.474785173667371,
+      "learning_rate": 7.591763468774139e-06,
+      "loss": 0.3091,
+      "step": 984
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.292113949232025,
+      "learning_rate": 7.586877076514266e-06,
+      "loss": 0.2832,
+      "step": 985
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.259429538183628,
+      "learning_rate": 7.5819873081948105e-06,
+      "loss": 0.27,
+      "step": 986
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.171249926692697,
+      "learning_rate": 7.577094170197268e-06,
+      "loss": 0.3425,
+      "step": 987
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.496084358698798,
+      "learning_rate": 7.572197668907533e-06,
+      "loss": 0.2871,
+      "step": 988
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.072296982688071,
+      "learning_rate": 7.567297810715893e-06,
+      "loss": 0.2808,
+      "step": 989
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.8890804218551915,
+      "learning_rate": 7.562394602017009e-06,
+      "loss": 0.2563,
+      "step": 990
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.985705361668515,
+      "learning_rate": 7.557488049209921e-06,
+      "loss": 0.2227,
+      "step": 991
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.390515717442826,
+      "learning_rate": 7.552578158698033e-06,
+      "loss": 0.3027,
+      "step": 992
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.189990838683105,
+      "learning_rate": 7.5476649368891004e-06,
+      "loss": 0.332,
+      "step": 993
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.440702295376607,
+      "learning_rate": 7.542748390195231e-06,
+      "loss": 0.46,
+      "step": 994
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.284920721689688,
+      "learning_rate": 7.537828525032871e-06,
+      "loss": 0.2966,
+      "step": 995
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.113087983464492,
+      "learning_rate": 7.532905347822792e-06,
+      "loss": 0.302,
+      "step": 996
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.065636498679776,
+      "learning_rate": 7.527978864990096e-06,
+      "loss": 0.2803,
+      "step": 997
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.9415966061480825,
+      "learning_rate": 7.523049082964195e-06,
+      "loss": 0.3408,
+      "step": 998
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.815399390144828,
+      "learning_rate": 7.518116008178805e-06,
+      "loss": 0.2854,
+      "step": 999
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.845868105874793,
+      "learning_rate": 7.5131796470719435e-06,
+      "loss": 0.2517,
+      "step": 1000
+    },
+    {
+      "epoch": 1.41,
+      "eval_avg_AUC": 0.7055294794561394,
+      "eval_avg_Accuracy": 0.6349883952254642,
+      "eval_avg_Accuracy-right": 0.970392591626451,
+      "eval_avg_Accuracy-wrong": 0.05014782806458949,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6232751879750892,
+      "eval_last_AUC": 0.730734067321652,
+      "eval_last_Accuracy": 0.6959549071618037,
+      "eval_last_Accuracy-right": 0.788183122472936,
+      "eval_last_Accuracy-wrong": 0.5351375938139641,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6361134787001135,
+      "eval_max_AUC": 0.5289832129694414,
+      "eval_max_Accuracy": 0.633289124668435,
+      "eval_max_Accuracy-right": 0.9930872570757793,
+      "eval_max_Accuracy-wrong": 0.005913122583579713,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5188094998445936,
+      "eval_min_AUC": 0.7317193529185435,
+      "eval_min_Accuracy": 0.6971568302387268,
+      "eval_min_Accuracy-right": 0.7815964523281597,
+      "eval_min_Accuracy-wrong": 0.5499204002729133,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6353902007845607,
+      "eval_prod_AUC": 0.7262823396045995,
+      "eval_prod_Accuracy": 0.630222148541114,
+      "eval_prod_Accuracy-right": 0.5374983696361028,
+      "eval_prod_Accuracy-wrong": 0.7919035706163293,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6234523293062538,
+      "eval_runtime": 667.6975,
+      "eval_samples_per_second": 36.136,
+      "eval_steps_per_second": 1.129,
+      "eval_sum_AUC": 0.5284847673065445,
+      "eval_sum_Accuracy": 0.6361488726790451,
+      "eval_sum_Accuracy-right": 0.9998695708882223,
+      "eval_sum_Accuracy-wrong": 0.001933136229247214,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.5944860468487428,
+      "step": 1000
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.856875713496743,
+      "learning_rate": 7.508240006085914e-06,
+      "loss": 0.2405,
+      "step": 1001
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.648116207929127,
+      "learning_rate": 7.5032970916673005e-06,
+      "loss": 0.2688,
+      "step": 1002
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.873820910680983,
+      "learning_rate": 7.49835091026696e-06,
+      "loss": 0.5376,
+      "step": 1003
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.147283066439303,
+      "learning_rate": 7.493401468340016e-06,
+      "loss": 0.2695,
+      "step": 1004
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.040340602351444,
+      "learning_rate": 7.48844877234584e-06,
+      "loss": 0.2697,
+      "step": 1005
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.353161380938701,
+      "learning_rate": 7.4834928287480566e-06,
+      "loss": 0.2812,
+      "step": 1006
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.149359695039955,
+      "learning_rate": 7.478533644014525e-06,
+      "loss": 0.313,
+      "step": 1007
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.767429302547887,
+      "learning_rate": 7.473571224617339e-06,
+      "loss": 0.2524,
+      "step": 1008
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.054109283382482,
+      "learning_rate": 7.468605577032808e-06,
+      "loss": 0.3062,
+      "step": 1009
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.956580153833473,
+      "learning_rate": 7.463636707741458e-06,
+      "loss": 0.2329,
+      "step": 1010
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.7535780737058975,
+      "learning_rate": 7.45866462322802e-06,
+      "loss": 0.3726,
+      "step": 1011
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.015231910832732,
+      "learning_rate": 7.453689329981417e-06,
+      "loss": 0.3501,
+      "step": 1012
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.265392366786086,
+      "learning_rate": 7.448710834494766e-06,
+      "loss": 0.3894,
+      "step": 1013
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.783435303720241,
+      "learning_rate": 7.443729143265357e-06,
+      "loss": 0.3203,
+      "step": 1014
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.671594938204128,
+      "learning_rate": 7.4387442627946536e-06,
+      "loss": 0.2954,
+      "step": 1015
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.024547554807496,
+      "learning_rate": 7.433756199588282e-06,
+      "loss": 0.2571,
+      "step": 1016
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.1029969019925545,
+      "learning_rate": 7.428764960156023e-06,
+      "loss": 0.3193,
+      "step": 1017
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.846023518048394,
+      "learning_rate": 7.423770551011797e-06,
+      "loss": 0.2891,
+      "step": 1018
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.64532243217749,
+      "learning_rate": 7.418772978673667e-06,
+      "loss": 0.4346,
+      "step": 1019
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.335509370589314,
+      "learning_rate": 7.413772249663825e-06,
+      "loss": 0.3259,
+      "step": 1020
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.474292348872879,
+      "learning_rate": 7.408768370508577e-06,
+      "loss": 0.3516,
+      "step": 1021
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.370223561099294,
+      "learning_rate": 7.403761347738343e-06,
+      "loss": 0.3032,
+      "step": 1022
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.135329251275284,
+      "learning_rate": 7.3987511878876485e-06,
+      "loss": 0.2832,
+      "step": 1023
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.429227120495372,
+      "learning_rate": 7.393737897495106e-06,
+      "loss": 0.2747,
+      "step": 1024
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.211424843874529,
+      "learning_rate": 7.388721483103423e-06,
+      "loss": 0.4419,
+      "step": 1025
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.617902364001967,
+      "learning_rate": 7.383701951259375e-06,
+      "loss": 0.2905,
+      "step": 1026
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.167835380885159,
+      "learning_rate": 7.378679308513812e-06,
+      "loss": 0.3638,
+      "step": 1027
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.364924967578536,
+      "learning_rate": 7.373653561421642e-06,
+      "loss": 0.3462,
+      "step": 1028
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.205687105986581,
+      "learning_rate": 7.368624716541824e-06,
+      "loss": 0.377,
+      "step": 1029
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.92428101245482,
+      "learning_rate": 7.3635927804373584e-06,
+      "loss": 0.304,
+      "step": 1030
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.975959103393626,
+      "learning_rate": 7.358557759675284e-06,
+      "loss": 0.3457,
+      "step": 1031
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.41950830757576,
+      "learning_rate": 7.353519660826665e-06,
+      "loss": 0.3528,
+      "step": 1032
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.583115106301743,
+      "learning_rate": 7.348478490466575e-06,
+      "loss": 0.3701,
+      "step": 1033
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.033915219169506,
+      "learning_rate": 7.343434255174105e-06,
+      "loss": 0.3071,
+      "step": 1034
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.66164893251374,
+      "learning_rate": 7.3383869615323445e-06,
+      "loss": 0.355,
+      "step": 1035
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.947922781399205,
+      "learning_rate": 7.333336616128369e-06,
+      "loss": 0.2917,
+      "step": 1036
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.189766252766152,
+      "learning_rate": 7.328283225553243e-06,
+      "loss": 0.2375,
+      "step": 1037
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.3986651000771175,
+      "learning_rate": 7.323226796402003e-06,
+      "loss": 0.2607,
+      "step": 1038
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.666018304326934,
+      "learning_rate": 7.318167335273647e-06,
+      "loss": 0.3218,
+      "step": 1039
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.410013671277181,
+      "learning_rate": 7.313104848771139e-06,
+      "loss": 0.2998,
+      "step": 1040
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.248329591044957,
+      "learning_rate": 7.308039343501381e-06,
+      "loss": 0.3682,
+      "step": 1041
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.3689395123049595,
+      "learning_rate": 7.302970826075222e-06,
+      "loss": 0.2727,
+      "step": 1042
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.859838426611454,
+      "learning_rate": 7.297899303107441e-06,
+      "loss": 0.4077,
+      "step": 1043
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.568557891619299,
+      "learning_rate": 7.2928247812167344e-06,
+      "loss": 0.2825,
+      "step": 1044
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.938811870692812,
+      "learning_rate": 7.287747267025719e-06,
+      "loss": 0.2798,
+      "step": 1045
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.311823129814038,
+      "learning_rate": 7.282666767160913e-06,
+      "loss": 0.2744,
+      "step": 1046
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.021142332121217,
+      "learning_rate": 7.277583288252733e-06,
+      "loss": 0.3591,
+      "step": 1047
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.585436598356468,
+      "learning_rate": 7.272496836935482e-06,
+      "loss": 0.3369,
+      "step": 1048
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.558383918352929,
+      "learning_rate": 7.267407419847341e-06,
+      "loss": 0.4385,
+      "step": 1049
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.759324513134668,
+      "learning_rate": 7.262315043630366e-06,
+      "loss": 0.2788,
+      "step": 1050
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.807028444596609,
+      "learning_rate": 7.2572197149304715e-06,
+      "loss": 0.3354,
+      "step": 1051
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.087769591714276,
+      "learning_rate": 7.252121440397424e-06,
+      "loss": 0.3491,
+      "step": 1052
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.848139377740482,
+      "learning_rate": 7.247020226684838e-06,
+      "loss": 0.3301,
+      "step": 1053
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 5.902424799339491,
+      "learning_rate": 7.241916080450163e-06,
+      "loss": 0.2227,
+      "step": 1054
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 5.255959767461447,
+      "learning_rate": 7.236809008354672e-06,
+      "loss": 0.2659,
+      "step": 1055
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.261449690863952,
+      "learning_rate": 7.23169901706346e-06,
+      "loss": 0.2628,
+      "step": 1056
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.165259538061054,
+      "learning_rate": 7.226586113245435e-06,
+      "loss": 0.3311,
+      "step": 1057
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.373773361094074,
+      "learning_rate": 7.221470303573298e-06,
+      "loss": 0.3823,
+      "step": 1058
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.537569060878586,
+      "learning_rate": 7.216351594723547e-06,
+      "loss": 0.2144,
+      "step": 1059
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.993217154099028,
+      "learning_rate": 7.211229993376465e-06,
+      "loss": 0.3125,
+      "step": 1060
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.635515442432376,
+      "learning_rate": 7.206105506216107e-06,
+      "loss": 0.2673,
+      "step": 1061
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.011511952154373,
+      "learning_rate": 7.200978139930294e-06,
+      "loss": 0.302,
+      "step": 1062
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.745479664430515,
+      "learning_rate": 7.195847901210608e-06,
+      "loss": 0.3525,
+      "step": 1063
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.2384409144691,
+      "learning_rate": 7.190714796752375e-06,
+      "loss": 0.2163,
+      "step": 1064
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.93530556668415,
+      "learning_rate": 7.185578833254665e-06,
+      "loss": 0.4316,
+      "step": 1065
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.6546924904276303,
+      "learning_rate": 7.180440017420277e-06,
+      "loss": 0.1711,
+      "step": 1066
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.544348211541799,
+      "learning_rate": 7.175298355955734e-06,
+      "loss": 0.2292,
+      "step": 1067
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.909795311028831,
+      "learning_rate": 7.170153855571272e-06,
+      "loss": 0.3074,
+      "step": 1068
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.034886708790083,
+      "learning_rate": 7.165006522980828e-06,
+      "loss": 0.2458,
+      "step": 1069
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.607206545379131,
+      "learning_rate": 7.1598563649020456e-06,
+      "loss": 0.2734,
+      "step": 1070
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.919364306341351,
+      "learning_rate": 7.154703388056246e-06,
+      "loss": 0.2388,
+      "step": 1071
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.117545310161974,
+      "learning_rate": 7.14954759916843e-06,
+      "loss": 0.2734,
+      "step": 1072
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.576331124673292,
+      "learning_rate": 7.144389004967276e-06,
+      "loss": 0.3901,
+      "step": 1073
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.7522801597987705,
+      "learning_rate": 7.139227612185115e-06,
+      "loss": 0.2383,
+      "step": 1074
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.591301221091546,
+      "learning_rate": 7.134063427557934e-06,
+      "loss": 0.2815,
+      "step": 1075
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.367510825647729,
+      "learning_rate": 7.128896457825364e-06,
+      "loss": 0.3079,
+      "step": 1076
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.692142265583918,
+      "learning_rate": 7.1237267097306696e-06,
+      "loss": 0.2576,
+      "step": 1077
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.531603105796802,
+      "learning_rate": 7.118554190020742e-06,
+      "loss": 0.2307,
+      "step": 1078
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.028550561887652,
+      "learning_rate": 7.11337890544609e-06,
+      "loss": 0.2864,
+      "step": 1079
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.59915497559429,
+      "learning_rate": 7.108200862760828e-06,
+      "loss": 0.2866,
+      "step": 1080
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.34154382123477,
+      "learning_rate": 7.103020068722675e-06,
+      "loss": 0.3557,
+      "step": 1081
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.857543232517111,
+      "learning_rate": 7.097836530092937e-06,
+      "loss": 0.4612,
+      "step": 1082
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.516548501876205,
+      "learning_rate": 7.092650253636498e-06,
+      "loss": 0.3677,
+      "step": 1083
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.64710553960146,
+      "learning_rate": 7.087461246121826e-06,
+      "loss": 0.2595,
+      "step": 1084
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.274678281968768,
+      "learning_rate": 7.082269514320944e-06,
+      "loss": 0.3667,
+      "step": 1085
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.965275172002132,
+      "learning_rate": 7.0770750650094335e-06,
+      "loss": 0.2771,
+      "step": 1086
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.4858655863472245,
+      "learning_rate": 7.071877904966422e-06,
+      "loss": 0.311,
+      "step": 1087
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.494666537297103,
+      "learning_rate": 7.066678040974575e-06,
+      "loss": 0.2383,
+      "step": 1088
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.704968631822675,
+      "learning_rate": 7.061475479820088e-06,
+      "loss": 0.4116,
+      "step": 1089
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.622580208886237,
+      "learning_rate": 7.056270228292674e-06,
+      "loss": 0.2688,
+      "step": 1090
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.017840310528078,
+      "learning_rate": 7.05106229318556e-06,
+      "loss": 0.3289,
+      "step": 1091
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.676018756940083,
+      "learning_rate": 7.045851681295472e-06,
+      "loss": 0.345,
+      "step": 1092
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.488139166146642,
+      "learning_rate": 7.040638399422632e-06,
+      "loss": 0.3379,
+      "step": 1093
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.064451541701058,
+      "learning_rate": 7.035422454370746e-06,
+      "loss": 0.2908,
+      "step": 1094
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.277416260461109,
+      "learning_rate": 7.030203852946995e-06,
+      "loss": 0.3108,
+      "step": 1095
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.201089466345504,
+      "learning_rate": 7.024982601962027e-06,
+      "loss": 0.2444,
+      "step": 1096
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.263281910819227,
+      "learning_rate": 7.019758708229949e-06,
+      "loss": 0.2952,
+      "step": 1097
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.87329000824749,
+      "learning_rate": 7.014532178568314e-06,
+      "loss": 0.3857,
+      "step": 1098
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.968658360877729,
+      "learning_rate": 7.0093030197981185e-06,
+      "loss": 0.2971,
+      "step": 1099
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.050236558790991,
+      "learning_rate": 7.004071238743787e-06,
+      "loss": 0.2866,
+      "step": 1100
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.31448770542556,
+      "learning_rate": 6.99883684223317e-06,
+      "loss": 0.3296,
+      "step": 1101
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.625960767080848,
+      "learning_rate": 6.993599837097527e-06,
+      "loss": 0.2146,
+      "step": 1102
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.276502955391807,
+      "learning_rate": 6.988360230171527e-06,
+      "loss": 0.3081,
+      "step": 1103
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.721541245826979,
+      "learning_rate": 6.98311802829323e-06,
+      "loss": 0.3633,
+      "step": 1104
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.442648505903865,
+      "learning_rate": 6.9778732383040825e-06,
+      "loss": 0.4282,
+      "step": 1105
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.892091946817631,
+      "learning_rate": 6.972625867048914e-06,
+      "loss": 0.2739,
+      "step": 1106
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.028513901338223,
+      "learning_rate": 6.96737592137592e-06,
+      "loss": 0.3208,
+      "step": 1107
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.56252203724774,
+      "learning_rate": 6.962123408136649e-06,
+      "loss": 0.3521,
+      "step": 1108
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.794561461369359,
+      "learning_rate": 6.9568683341860135e-06,
+      "loss": 0.2385,
+      "step": 1109
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.535610473669975,
+      "learning_rate": 6.951610706382257e-06,
+      "loss": 0.2922,
+      "step": 1110
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.19374551397145,
+      "learning_rate": 6.946350531586959e-06,
+      "loss": 0.3599,
+      "step": 1111
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.250630515285008,
+      "learning_rate": 6.941087816665025e-06,
+      "loss": 0.2954,
+      "step": 1112
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.591136960367386,
+      "learning_rate": 6.935822568484676e-06,
+      "loss": 0.241,
+      "step": 1113
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.289501602632274,
+      "learning_rate": 6.930554793917434e-06,
+      "loss": 0.3313,
+      "step": 1114
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.858319690592284,
+      "learning_rate": 6.925284499838125e-06,
+      "loss": 0.3513,
+      "step": 1115
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.915923377175696,
+      "learning_rate": 6.9200116931248575e-06,
+      "loss": 0.356,
+      "step": 1116
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.4454977080697216,
+      "learning_rate": 6.914736380659023e-06,
+      "loss": 0.3167,
+      "step": 1117
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.390884690879619,
+      "learning_rate": 6.909458569325281e-06,
+      "loss": 0.3123,
+      "step": 1118
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.03427189638723,
+      "learning_rate": 6.90417826601155e-06,
+      "loss": 0.3062,
+      "step": 1119
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.054236855676072,
+      "learning_rate": 6.898895477609007e-06,
+      "loss": 0.2554,
+      "step": 1120
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.4355349861933835,
+      "learning_rate": 6.893610211012067e-06,
+      "loss": 0.3726,
+      "step": 1121
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.202786875426783,
+      "learning_rate": 6.888322473118381e-06,
+      "loss": 0.2891,
+      "step": 1122
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.48086992113035,
+      "learning_rate": 6.8830322708288255e-06,
+      "loss": 0.3291,
+      "step": 1123
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.879855272151002,
+      "learning_rate": 6.877739611047492e-06,
+      "loss": 0.365,
+      "step": 1124
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.353591720514249,
+      "learning_rate": 6.872444500681681e-06,
+      "loss": 0.3352,
+      "step": 1125
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.640500354813242,
+      "learning_rate": 6.8671469466418914e-06,
+      "loss": 0.2473,
+      "step": 1126
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.119019774956855,
+      "learning_rate": 6.861846955841808e-06,
+      "loss": 0.2192,
+      "step": 1127
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.358120527893062,
+      "learning_rate": 6.856544535198296e-06,
+      "loss": 0.3384,
+      "step": 1128
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.48880636995769,
+      "learning_rate": 6.8512396916313985e-06,
+      "loss": 0.2769,
+      "step": 1129
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.828168340969308,
+      "learning_rate": 6.845932432064312e-06,
+      "loss": 0.2761,
+      "step": 1130
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.360303744547078,
+      "learning_rate": 6.840622763423391e-06,
+      "loss": 0.3647,
+      "step": 1131
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.031080380543954,
+      "learning_rate": 6.835310692638132e-06,
+      "loss": 0.2766,
+      "step": 1132
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.691612773583568,
+      "learning_rate": 6.829996226641168e-06,
+      "loss": 0.1852,
+      "step": 1133
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.903070234671075,
+      "learning_rate": 6.824679372368258e-06,
+      "loss": 0.3882,
+      "step": 1134
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.1850707235668345,
+      "learning_rate": 6.819360136758277e-06,
+      "loss": 0.2412,
+      "step": 1135
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.641282193765888,
+      "learning_rate": 6.814038526753205e-06,
+      "loss": 0.2812,
+      "step": 1136
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.04340827290255,
+      "learning_rate": 6.8087145492981275e-06,
+      "loss": 0.2942,
+      "step": 1137
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.671024312961727,
+      "learning_rate": 6.803388211341213e-06,
+      "loss": 0.3528,
+      "step": 1138
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 16.23218928316647,
+      "learning_rate": 6.7980595198337155e-06,
+      "loss": 0.4172,
+      "step": 1139
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.644830722944572,
+      "learning_rate": 6.7927284817299566e-06,
+      "loss": 0.3901,
+      "step": 1140
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.409002051374585,
+      "learning_rate": 6.787395103987323e-06,
+      "loss": 0.332,
+      "step": 1141
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.939821765934187,
+      "learning_rate": 6.782059393566254e-06,
+      "loss": 0.3513,
+      "step": 1142
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.901175855443189,
+      "learning_rate": 6.7767213574302344e-06,
+      "loss": 0.3467,
+      "step": 1143
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.333944986805026,
+      "learning_rate": 6.771381002545782e-06,
+      "loss": 0.3696,
+      "step": 1144
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.153252931528158,
+      "learning_rate": 6.766038335882441e-06,
+      "loss": 0.2615,
+      "step": 1145
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.523887270554477,
+      "learning_rate": 6.760693364412776e-06,
+      "loss": 0.2764,
+      "step": 1146
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.180201503821454,
+      "learning_rate": 6.755346095112354e-06,
+      "loss": 0.4453,
+      "step": 1147
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.730138305517497,
+      "learning_rate": 6.749996534959746e-06,
+      "loss": 0.2795,
+      "step": 1148
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.822525920251833,
+      "learning_rate": 6.744644690936509e-06,
+      "loss": 0.3167,
+      "step": 1149
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.081020970325389,
+      "learning_rate": 6.739290570027185e-06,
+      "loss": 0.375,
+      "step": 1150
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.03451378016529,
+      "learning_rate": 6.733934179219281e-06,
+      "loss": 0.2212,
+      "step": 1151
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.7075728315717,
+      "learning_rate": 6.728575525503273e-06,
+      "loss": 0.2764,
+      "step": 1152
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.89819009256561,
+      "learning_rate": 6.723214615872585e-06,
+      "loss": 0.374,
+      "step": 1153
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.746039667290383,
+      "learning_rate": 6.7178514573235895e-06,
+      "loss": 0.4072,
+      "step": 1154
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.421054628822626,
+      "learning_rate": 6.712486056855591e-06,
+      "loss": 0.3396,
+      "step": 1155
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.082527316099785,
+      "learning_rate": 6.707118421470822e-06,
+      "loss": 0.3545,
+      "step": 1156
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.950129242263197,
+      "learning_rate": 6.701748558174428e-06,
+      "loss": 0.2644,
+      "step": 1157
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.949763764629544,
+      "learning_rate": 6.696376473974466e-06,
+      "loss": 0.3643,
+      "step": 1158
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.58668411438948,
+      "learning_rate": 6.691002175881891e-06,
+      "loss": 0.3203,
+      "step": 1159
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.8974463577091605,
+      "learning_rate": 6.685625670910545e-06,
+      "loss": 0.2446,
+      "step": 1160
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.791860215901977,
+      "learning_rate": 6.680246966077151e-06,
+      "loss": 0.2622,
+      "step": 1161
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.964224176065095,
+      "learning_rate": 6.674866068401306e-06,
+      "loss": 0.2288,
+      "step": 1162
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.584310217799718,
+      "learning_rate": 6.669482984905464e-06,
+      "loss": 0.3091,
+      "step": 1163
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.149226380824308,
+      "learning_rate": 6.664097722614934e-06,
+      "loss": 0.313,
+      "step": 1164
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.892128570925954,
+      "learning_rate": 6.658710288557871e-06,
+      "loss": 0.2507,
+      "step": 1165
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.287806782132015,
+      "learning_rate": 6.653320689765257e-06,
+      "loss": 0.2252,
+      "step": 1166
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.455345651654644,
+      "learning_rate": 6.647928933270908e-06,
+      "loss": 0.3853,
+      "step": 1167
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.238676690170254,
+      "learning_rate": 6.642535026111451e-06,
+      "loss": 0.3286,
+      "step": 1168
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.031357046728988,
+      "learning_rate": 6.6371389753263186e-06,
+      "loss": 0.2795,
+      "step": 1169
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.678953450420682,
+      "learning_rate": 6.631740787957745e-06,
+      "loss": 0.2732,
+      "step": 1170
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.093698914465941,
+      "learning_rate": 6.6263404710507495e-06,
+      "loss": 0.2322,
+      "step": 1171
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.793158772673536,
+      "learning_rate": 6.620938031653133e-06,
+      "loss": 0.3777,
+      "step": 1172
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.341164056671769,
+      "learning_rate": 6.615533476815465e-06,
+      "loss": 0.2715,
+      "step": 1173
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.381307540527043,
+      "learning_rate": 6.610126813591077e-06,
+      "loss": 0.2434,
+      "step": 1174
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.019031865723753,
+      "learning_rate": 6.604718049036047e-06,
+      "loss": 0.3291,
+      "step": 1175
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.880259324860161,
+      "learning_rate": 6.599307190209206e-06,
+      "loss": 0.2756,
+      "step": 1176
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.063248551089938,
+      "learning_rate": 6.5938942441721065e-06,
+      "loss": 0.4238,
+      "step": 1177
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.1981588879602,
+      "learning_rate": 6.588479217989031e-06,
+      "loss": 0.3457,
+      "step": 1178
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 15.247753283483625,
+      "learning_rate": 6.583062118726979e-06,
+      "loss": 0.4531,
+      "step": 1179
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.187556037165955,
+      "learning_rate": 6.577642953455649e-06,
+      "loss": 0.2156,
+      "step": 1180
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.6993099853645255,
+      "learning_rate": 6.572221729247441e-06,
+      "loss": 0.2688,
+      "step": 1181
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.451253372490162,
+      "learning_rate": 6.56679845317744e-06,
+      "loss": 0.2991,
+      "step": 1182
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.448780426257079,
+      "learning_rate": 6.561373132323407e-06,
+      "loss": 0.3044,
+      "step": 1183
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.734155197162429,
+      "learning_rate": 6.555945773765778e-06,
+      "loss": 0.3115,
+      "step": 1184
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.831453028145821,
+      "learning_rate": 6.550516384587642e-06,
+      "loss": 0.3188,
+      "step": 1185
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.478895617437194,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.4355,
+      "step": 1186
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.22812873159387,
+      "learning_rate": 6.539651542715451e-06,
+      "loss": 0.2764,
+      "step": 1187
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.666880586040801,
+      "learning_rate": 6.5342161042007926e-06,
+      "loss": 0.3164,
+      "step": 1188
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.97068509741892,
+      "learning_rate": 6.5287786634244e-06,
+      "loss": 0.282,
+      "step": 1189
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.422078799929475,
+      "learning_rate": 6.523339227482522e-06,
+      "loss": 0.2551,
+      "step": 1190
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.817504341138592,
+      "learning_rate": 6.517897803474011e-06,
+      "loss": 0.2986,
+      "step": 1191
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.24357608715862,
+      "learning_rate": 6.512454398500316e-06,
+      "loss": 0.3716,
+      "step": 1192
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.757262574890522,
+      "learning_rate": 6.507009019665469e-06,
+      "loss": 0.3442,
+      "step": 1193
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.437530216531816,
+      "learning_rate": 6.501561674076078e-06,
+      "loss": 0.3149,
+      "step": 1194
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.499307114679,
+      "learning_rate": 6.4961123688413175e-06,
+      "loss": 0.2363,
+      "step": 1195
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.780345514142386,
+      "learning_rate": 6.490661111072923e-06,
+      "loss": 0.3472,
+      "step": 1196
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.602614506934005,
+      "learning_rate": 6.485207907885175e-06,
+      "loss": 0.4207,
+      "step": 1197
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.399794942090772,
+      "learning_rate": 6.479752766394894e-06,
+      "loss": 0.416,
+      "step": 1198
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.52054286653332,
+      "learning_rate": 6.4742956937214285e-06,
+      "loss": 0.3677,
+      "step": 1199
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.874760156831451,
+      "learning_rate": 6.468836696986649e-06,
+      "loss": 0.2798,
+      "step": 1200
+    },
+    {
+      "epoch": 1.69,
+      "eval_avg_AUC": 0.7437681536959805,
+      "eval_avg_Accuracy": 0.6627984084880637,
+      "eval_avg_Accuracy-right": 0.9205034563714621,
+      "eval_avg_Accuracy-wrong": 0.21344098248806004,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6409851021111105,
+      "eval_last_AUC": 0.7763218302292323,
+      "eval_last_Accuracy": 0.7123259283819628,
+      "eval_last_Accuracy-right": 0.7504238946132776,
+      "eval_last_Accuracy-wrong": 0.6458949283602456,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6506704634912995,
+      "eval_max_AUC": 0.5477644080934393,
+      "eval_max_Accuracy": 0.6347811671087533,
+      "eval_max_Accuracy-right": 0.9838267901395592,
+      "eval_max_Accuracy-wrong": 0.026154196042756424,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5325709929872469,
+      "eval_min_AUC": 0.7707196596064185,
+      "eval_min_Accuracy": 0.7110411140583555,
+      "eval_min_Accuracy-right": 0.7375766271031694,
+      "eval_min_Accuracy-wrong": 0.6647714350693654,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6466059977289114,
+      "eval_prod_AUC": 0.7674564845975738,
+      "eval_prod_Accuracy": 0.6281084217506632,
+      "eval_prod_Accuracy-right": 0.49647841398200077,
+      "eval_prod_Accuracy-wrong": 0.8576302024107346,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6460802189162478,
+      "eval_runtime": 668.2977,
+      "eval_samples_per_second": 36.104,
+      "eval_steps_per_second": 1.128,
+      "eval_sum_AUC": 0.5833287478962483,
+      "eval_sum_Accuracy": 0.6361074270557029,
+      "eval_sum_Accuracy-right": 0.9964784139820008,
+      "eval_sum_Accuracy-wrong": 0.007732544916988856,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.619962648462929,
+      "step": 1200
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.700404627909396,
+      "learning_rate": 6.463375783314938e-06,
+      "loss": 0.4043,
+      "step": 1201
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.882760175710391,
+      "learning_rate": 6.457912959833177e-06,
+      "loss": 0.2656,
+      "step": 1202
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.44968763366824,
+      "learning_rate": 6.45244823367074e-06,
+      "loss": 0.3457,
+      "step": 1203
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.108947044790602,
+      "learning_rate": 6.44698161195949e-06,
+      "loss": 0.3345,
+      "step": 1204
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.962436375870858,
+      "learning_rate": 6.441513101833755e-06,
+      "loss": 0.2241,
+      "step": 1205
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.28700721543711,
+      "learning_rate": 6.4360427104303326e-06,
+      "loss": 0.3003,
+      "step": 1206
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.246628810310547,
+      "learning_rate": 6.4305704448884756e-06,
+      "loss": 0.2937,
+      "step": 1207
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.669792211708034,
+      "learning_rate": 6.425096312349881e-06,
+      "loss": 0.238,
+      "step": 1208
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.43959056644161,
+      "learning_rate": 6.419620319958685e-06,
+      "loss": 0.3865,
+      "step": 1209
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.64480085647091,
+      "learning_rate": 6.414142474861448e-06,
+      "loss": 0.2712,
+      "step": 1210
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.717875100935446,
+      "learning_rate": 6.408662784207149e-06,
+      "loss": 0.4702,
+      "step": 1211
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.317406768091054,
+      "learning_rate": 6.403181255147177e-06,
+      "loss": 0.2625,
+      "step": 1212
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.845395346575547,
+      "learning_rate": 6.397697894835321e-06,
+      "loss": 0.3174,
+      "step": 1213
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.902102013023462,
+      "learning_rate": 6.392212710427754e-06,
+      "loss": 0.3232,
+      "step": 1214
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.685460711617788,
+      "learning_rate": 6.386725709083039e-06,
+      "loss": 0.3296,
+      "step": 1215
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.616690295940373,
+      "learning_rate": 6.381236897962102e-06,
+      "loss": 0.3926,
+      "step": 1216
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.869253598368983,
+      "learning_rate": 6.375746284228235e-06,
+      "loss": 0.2939,
+      "step": 1217
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 5.9921488587795695,
+      "learning_rate": 6.370253875047081e-06,
+      "loss": 0.2339,
+      "step": 1218
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.612289074477548,
+      "learning_rate": 6.364759677586627e-06,
+      "loss": 0.2766,
+      "step": 1219
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.195486706251973,
+      "learning_rate": 6.359263699017194e-06,
+      "loss": 0.4785,
+      "step": 1220
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.758113759024981,
+      "learning_rate": 6.3537659465114275e-06,
+      "loss": 0.3838,
+      "step": 1221
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.531065580964558,
+      "learning_rate": 6.348266427244286e-06,
+      "loss": 0.374,
+      "step": 1222
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.899299855330129,
+      "learning_rate": 6.3427651483930375e-06,
+      "loss": 0.2634,
+      "step": 1223
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.204200421440772,
+      "learning_rate": 6.337262117137243e-06,
+      "loss": 0.3091,
+      "step": 1224
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.477863889685757,
+      "learning_rate": 6.331757340658751e-06,
+      "loss": 0.2993,
+      "step": 1225
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.8069607487207096,
+      "learning_rate": 6.326250826141689e-06,
+      "loss": 0.3501,
+      "step": 1226
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.66326501313533,
+      "learning_rate": 6.320742580772455e-06,
+      "loss": 0.259,
+      "step": 1227
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.620801195058856,
+      "learning_rate": 6.3152326117397e-06,
+      "loss": 0.2871,
+      "step": 1228
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.524999692158317,
+      "learning_rate": 6.309720926234328e-06,
+      "loss": 0.2605,
+      "step": 1229
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.26463843832759,
+      "learning_rate": 6.304207531449486e-06,
+      "loss": 0.3086,
+      "step": 1230
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.7326019188202695,
+      "learning_rate": 6.298692434580543e-06,
+      "loss": 0.2262,
+      "step": 1231
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.959486406078255,
+      "learning_rate": 6.293175642825099e-06,
+      "loss": 0.2605,
+      "step": 1232
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.357126070263353,
+      "learning_rate": 6.287657163382961e-06,
+      "loss": 0.3169,
+      "step": 1233
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.90477185624728,
+      "learning_rate": 6.282137003456137e-06,
+      "loss": 0.3264,
+      "step": 1234
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.887481785985506,
+      "learning_rate": 6.276615170248833e-06,
+      "loss": 0.3254,
+      "step": 1235
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.698823233175139,
+      "learning_rate": 6.271091670967437e-06,
+      "loss": 0.2883,
+      "step": 1236
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.3093794251897055,
+      "learning_rate": 6.265566512820506e-06,
+      "loss": 0.3467,
+      "step": 1237
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.8876440932638445,
+      "learning_rate": 6.260039703018771e-06,
+      "loss": 0.2571,
+      "step": 1238
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.142654536987898,
+      "learning_rate": 6.254511248775112e-06,
+      "loss": 0.2747,
+      "step": 1239
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.958296760075168,
+      "learning_rate": 6.248981157304559e-06,
+      "loss": 0.2742,
+      "step": 1240
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.910454462480344,
+      "learning_rate": 6.243449435824276e-06,
+      "loss": 0.2537,
+      "step": 1241
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.724243034213231,
+      "learning_rate": 6.237916091553552e-06,
+      "loss": 0.2581,
+      "step": 1242
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.1323404689424335,
+      "learning_rate": 6.2323811317138015e-06,
+      "loss": 0.251,
+      "step": 1243
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.927939223067512,
+      "learning_rate": 6.22684456352854e-06,
+      "loss": 0.3042,
+      "step": 1244
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.506658921046029,
+      "learning_rate": 6.221306394223387e-06,
+      "loss": 0.3906,
+      "step": 1245
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.42426219035878,
+      "learning_rate": 6.215766631026049e-06,
+      "loss": 0.3486,
+      "step": 1246
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.675998764505445,
+      "learning_rate": 6.210225281166312e-06,
+      "loss": 0.261,
+      "step": 1247
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.799661816234979,
+      "learning_rate": 6.204682351876035e-06,
+      "loss": 0.3472,
+      "step": 1248
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.13167253385321,
+      "learning_rate": 6.199137850389137e-06,
+      "loss": 0.3843,
+      "step": 1249
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.7904328041989075,
+      "learning_rate": 6.193591783941586e-06,
+      "loss": 0.3342,
+      "step": 1250
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.497034704085983,
+      "learning_rate": 6.1880441597714e-06,
+      "loss": 0.2739,
+      "step": 1251
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.13613384868159,
+      "learning_rate": 6.182494985118625e-06,
+      "loss": 0.3252,
+      "step": 1252
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.43074437393179,
+      "learning_rate": 6.176944267225326e-06,
+      "loss": 0.3604,
+      "step": 1253
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.83970958314763,
+      "learning_rate": 6.171392013335591e-06,
+      "loss": 0.3252,
+      "step": 1254
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.24789699997936,
+      "learning_rate": 6.1658382306955075e-06,
+      "loss": 0.2285,
+      "step": 1255
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.433225047177424,
+      "learning_rate": 6.1602829265531585e-06,
+      "loss": 0.2651,
+      "step": 1256
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.395434932021952,
+      "learning_rate": 6.154726108158613e-06,
+      "loss": 0.2996,
+      "step": 1257
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.954862212707054,
+      "learning_rate": 6.1491677827639176e-06,
+      "loss": 0.3596,
+      "step": 1258
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.558686626159908,
+      "learning_rate": 6.143607957623082e-06,
+      "loss": 0.2651,
+      "step": 1259
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.236121179769956,
+      "learning_rate": 6.138046639992079e-06,
+      "loss": 0.3726,
+      "step": 1260
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.593818863960344,
+      "learning_rate": 6.132483837128823e-06,
+      "loss": 0.3799,
+      "step": 1261
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.623881869246551,
+      "learning_rate": 6.126919556293171e-06,
+      "loss": 0.2791,
+      "step": 1262
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.149602088071063,
+      "learning_rate": 6.121353804746907e-06,
+      "loss": 0.2942,
+      "step": 1263
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.279102007071229,
+      "learning_rate": 6.115786589753735e-06,
+      "loss": 0.2725,
+      "step": 1264
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.750028758457928,
+      "learning_rate": 6.1102179185792676e-06,
+      "loss": 0.3115,
+      "step": 1265
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.256092592519597,
+      "learning_rate": 6.1046477984910215e-06,
+      "loss": 0.3291,
+      "step": 1266
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.963803778969993,
+      "learning_rate": 6.099076236758398e-06,
+      "loss": 0.2961,
+      "step": 1267
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.668810149924438,
+      "learning_rate": 6.093503240652687e-06,
+      "loss": 0.4185,
+      "step": 1268
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.647316951951765,
+      "learning_rate": 6.0879288174470475e-06,
+      "loss": 0.3047,
+      "step": 1269
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.315310160699287,
+      "learning_rate": 6.082352974416496e-06,
+      "loss": 0.3311,
+      "step": 1270
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.54313595607463,
+      "learning_rate": 6.076775718837911e-06,
+      "loss": 0.4023,
+      "step": 1271
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.488249311127344,
+      "learning_rate": 6.071197057990007e-06,
+      "loss": 0.3169,
+      "step": 1272
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.101264623003464,
+      "learning_rate": 6.065616999153335e-06,
+      "loss": 0.2527,
+      "step": 1273
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.703401826648744,
+      "learning_rate": 6.060035549610275e-06,
+      "loss": 0.2463,
+      "step": 1274
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.216445545678997,
+      "learning_rate": 6.054452716645013e-06,
+      "loss": 0.3274,
+      "step": 1275
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.498746506650453,
+      "learning_rate": 6.048868507543547e-06,
+      "loss": 0.2988,
+      "step": 1276
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.491442937503534,
+      "learning_rate": 6.04328292959367e-06,
+      "loss": 0.3628,
+      "step": 1277
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.762553988221821,
+      "learning_rate": 6.03769599008496e-06,
+      "loss": 0.2351,
+      "step": 1278
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.765790643585311,
+      "learning_rate": 6.032107696308773e-06,
+      "loss": 0.2561,
+      "step": 1279
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.5135520589752725,
+      "learning_rate": 6.026518055558233e-06,
+      "loss": 0.2952,
+      "step": 1280
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.629900314057254,
+      "learning_rate": 6.0209270751282165e-06,
+      "loss": 0.3657,
+      "step": 1281
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.9480577183818,
+      "learning_rate": 6.015334762315358e-06,
+      "loss": 0.3228,
+      "step": 1282
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.477004021348314,
+      "learning_rate": 6.0097411244180225e-06,
+      "loss": 0.3789,
+      "step": 1283
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.022076093100977,
+      "learning_rate": 6.004146168736309e-06,
+      "loss": 0.2944,
+      "step": 1284
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.9798709520635835,
+      "learning_rate": 5.9985499025720354e-06,
+      "loss": 0.3367,
+      "step": 1285
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.111879836106336,
+      "learning_rate": 5.9929523332287275e-06,
+      "loss": 0.1736,
+      "step": 1286
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.262965177145975,
+      "learning_rate": 5.987353468011614e-06,
+      "loss": 0.3403,
+      "step": 1287
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.034549853391633,
+      "learning_rate": 5.981753314227616e-06,
+      "loss": 0.3157,
+      "step": 1288
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.598251149594113,
+      "learning_rate": 5.9761518791853334e-06,
+      "loss": 0.2151,
+      "step": 1289
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.33065189870276,
+      "learning_rate": 5.970549170195038e-06,
+      "loss": 0.2588,
+      "step": 1290
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.00363506648489,
+      "learning_rate": 5.964945194568669e-06,
+      "loss": 0.4502,
+      "step": 1291
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.337790874292512,
+      "learning_rate": 5.959339959619812e-06,
+      "loss": 0.3755,
+      "step": 1292
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.857603513928837,
+      "learning_rate": 5.953733472663702e-06,
+      "loss": 0.2878,
+      "step": 1293
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.634949703947829,
+      "learning_rate": 5.948125741017204e-06,
+      "loss": 0.4868,
+      "step": 1294
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.598496536368266,
+      "learning_rate": 5.94251677199881e-06,
+      "loss": 0.3394,
+      "step": 1295
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.30004840112656,
+      "learning_rate": 5.936906572928625e-06,
+      "loss": 0.2979,
+      "step": 1296
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 16.063062246179914,
+      "learning_rate": 5.931295151128359e-06,
+      "loss": 0.3877,
+      "step": 1297
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 12.302436295853953,
+      "learning_rate": 5.9256825139213206e-06,
+      "loss": 0.2979,
+      "step": 1298
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.725994787163433,
+      "learning_rate": 5.920068668632401e-06,
+      "loss": 0.2861,
+      "step": 1299
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.740361098676854,
+      "learning_rate": 5.914453622588071e-06,
+      "loss": 0.2844,
+      "step": 1300
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.232214196701149,
+      "learning_rate": 5.908837383116367e-06,
+      "loss": 0.3572,
+      "step": 1301
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.002339454928821,
+      "learning_rate": 5.9032199575468825e-06,
+      "loss": 0.3066,
+      "step": 1302
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 14.90714565322871,
+      "learning_rate": 5.897601353210762e-06,
+      "loss": 0.4473,
+      "step": 1303
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.3266075561130615,
+      "learning_rate": 5.891981577440682e-06,
+      "loss": 0.2815,
+      "step": 1304
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.876652992131552,
+      "learning_rate": 5.886360637570856e-06,
+      "loss": 0.3826,
+      "step": 1305
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.7976311590658725,
+      "learning_rate": 5.880738540937008e-06,
+      "loss": 0.2964,
+      "step": 1306
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.032546025889645,
+      "learning_rate": 5.8751152948763815e-06,
+      "loss": 0.2673,
+      "step": 1307
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.087598705262215,
+      "learning_rate": 5.86949090672771e-06,
+      "loss": 0.3091,
+      "step": 1308
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.005224251865695,
+      "learning_rate": 5.863865383831223e-06,
+      "loss": 0.207,
+      "step": 1309
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.943988218221327,
+      "learning_rate": 5.858238733528634e-06,
+      "loss": 0.2424,
+      "step": 1310
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.151525477801975,
+      "learning_rate": 5.85261096316312e-06,
+      "loss": 0.2485,
+      "step": 1311
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.683218790340961,
+      "learning_rate": 5.846982080079324e-06,
+      "loss": 0.2776,
+      "step": 1312
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 14.95570778333526,
+      "learning_rate": 5.841352091623345e-06,
+      "loss": 0.2764,
+      "step": 1313
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.079559719825116,
+      "learning_rate": 5.835721005142715e-06,
+      "loss": 0.2437,
+      "step": 1314
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.644040820976744,
+      "learning_rate": 5.830088827986408e-06,
+      "loss": 0.2053,
+      "step": 1315
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.333673928179607,
+      "learning_rate": 5.824455567504817e-06,
+      "loss": 0.3184,
+      "step": 1316
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.108462607186791,
+      "learning_rate": 5.81882123104975e-06,
+      "loss": 0.3423,
+      "step": 1317
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.74946145225459,
+      "learning_rate": 5.813185825974419e-06,
+      "loss": 0.2432,
+      "step": 1318
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.664144279315488,
+      "learning_rate": 5.80754935963343e-06,
+      "loss": 0.4487,
+      "step": 1319
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.96924902863016,
+      "learning_rate": 5.8019118393827765e-06,
+      "loss": 0.4045,
+      "step": 1320
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.665898016929757,
+      "learning_rate": 5.796273272579823e-06,
+      "loss": 0.3833,
+      "step": 1321
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.542228634406015,
+      "learning_rate": 5.790633666583305e-06,
+      "loss": 0.3638,
+      "step": 1322
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.877636250439117,
+      "learning_rate": 5.78499302875331e-06,
+      "loss": 0.3066,
+      "step": 1323
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.772859221753492,
+      "learning_rate": 5.779351366451274e-06,
+      "loss": 0.3101,
+      "step": 1324
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.981037854093982,
+      "learning_rate": 5.773708687039971e-06,
+      "loss": 0.3506,
+      "step": 1325
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.825173784923526,
+      "learning_rate": 5.7680649978834976e-06,
+      "loss": 0.2595,
+      "step": 1326
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.95934530326654,
+      "learning_rate": 5.762420306347276e-06,
+      "loss": 0.3198,
+      "step": 1327
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.177194813415156,
+      "learning_rate": 5.756774619798031e-06,
+      "loss": 0.3188,
+      "step": 1328
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.098526956897186,
+      "learning_rate": 5.751127945603786e-06,
+      "loss": 0.2917,
+      "step": 1329
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.15165580387882,
+      "learning_rate": 5.745480291133856e-06,
+      "loss": 0.3835,
+      "step": 1330
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.079073630189823,
+      "learning_rate": 5.739831663758834e-06,
+      "loss": 0.3833,
+      "step": 1331
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.21028451231747,
+      "learning_rate": 5.73418207085058e-06,
+      "loss": 0.303,
+      "step": 1332
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.376793135162231,
+      "learning_rate": 5.728531519782221e-06,
+      "loss": 0.2554,
+      "step": 1333
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.134455885849064,
+      "learning_rate": 5.722880017928128e-06,
+      "loss": 0.3872,
+      "step": 1334
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.935418367884691,
+      "learning_rate": 5.717227572663915e-06,
+      "loss": 0.2856,
+      "step": 1335
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.212768295363841,
+      "learning_rate": 5.711574191366427e-06,
+      "loss": 0.3088,
+      "step": 1336
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.402593567064902,
+      "learning_rate": 5.705919881413732e-06,
+      "loss": 0.2925,
+      "step": 1337
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.805871959423523,
+      "learning_rate": 5.700264650185108e-06,
+      "loss": 0.3247,
+      "step": 1338
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.956673258397387,
+      "learning_rate": 5.694608505061036e-06,
+      "loss": 0.3076,
+      "step": 1339
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.083452641304675,
+      "learning_rate": 5.68895145342319e-06,
+      "loss": 0.2886,
+      "step": 1340
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.2896894502125775,
+      "learning_rate": 5.683293502654429e-06,
+      "loss": 0.2908,
+      "step": 1341
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.870178706562168,
+      "learning_rate": 5.6776346601387826e-06,
+      "loss": 0.2627,
+      "step": 1342
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.5822304667322795,
+      "learning_rate": 5.671974933261441e-06,
+      "loss": 0.218,
+      "step": 1343
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.869577924005488,
+      "learning_rate": 5.666314329408758e-06,
+      "loss": 0.3009,
+      "step": 1344
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.790309426650434,
+      "learning_rate": 5.660652855968223e-06,
+      "loss": 0.2435,
+      "step": 1345
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.07571312017229,
+      "learning_rate": 5.654990520328465e-06,
+      "loss": 0.3542,
+      "step": 1346
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.703333671565709,
+      "learning_rate": 5.649327329879236e-06,
+      "loss": 0.2891,
+      "step": 1347
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.689069059233105,
+      "learning_rate": 5.643663292011405e-06,
+      "loss": 0.3167,
+      "step": 1348
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.145138402153409,
+      "learning_rate": 5.637998414116949e-06,
+      "loss": 0.3291,
+      "step": 1349
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.661210659501524,
+      "learning_rate": 5.632332703588935e-06,
+      "loss": 0.3862,
+      "step": 1350
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.804972113852745,
+      "learning_rate": 5.626666167821522e-06,
+      "loss": 0.2048,
+      "step": 1351
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.301517305418891,
+      "learning_rate": 5.620998814209945e-06,
+      "loss": 0.2559,
+      "step": 1352
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.321252546296352,
+      "learning_rate": 5.6153306501505045e-06,
+      "loss": 0.3179,
+      "step": 1353
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 14.897852892622762,
+      "learning_rate": 5.60966168304056e-06,
+      "loss": 0.2756,
+      "step": 1354
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.488788608214001,
+      "learning_rate": 5.60399192027852e-06,
+      "loss": 0.3037,
+      "step": 1355
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.549740287364084,
+      "learning_rate": 5.59832136926383e-06,
+      "loss": 0.2288,
+      "step": 1356
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.792203252766702,
+      "learning_rate": 5.592650037396965e-06,
+      "loss": 0.2629,
+      "step": 1357
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.981465799973394,
+      "learning_rate": 5.586977932079417e-06,
+      "loss": 0.272,
+      "step": 1358
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.508609191181147,
+      "learning_rate": 5.581305060713691e-06,
+      "loss": 0.2839,
+      "step": 1359
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.043818072182209,
+      "learning_rate": 5.575631430703289e-06,
+      "loss": 0.3425,
+      "step": 1360
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.749681284497589,
+      "learning_rate": 5.569957049452703e-06,
+      "loss": 0.2458,
+      "step": 1361
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.886485336897406,
+      "learning_rate": 5.5642819243674085e-06,
+      "loss": 0.321,
+      "step": 1362
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.47939742291272,
+      "learning_rate": 5.558606062853849e-06,
+      "loss": 0.4639,
+      "step": 1363
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.427318327819219,
+      "learning_rate": 5.552929472319428e-06,
+      "loss": 0.4277,
+      "step": 1364
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.011069076818455,
+      "learning_rate": 5.547252160172503e-06,
+      "loss": 0.2876,
+      "step": 1365
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 13.743010212306386,
+      "learning_rate": 5.541574133822374e-06,
+      "loss": 0.2693,
+      "step": 1366
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.796442350630697,
+      "learning_rate": 5.53589540067927e-06,
+      "loss": 0.3501,
+      "step": 1367
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.905867444491088,
+      "learning_rate": 5.5302159681543424e-06,
+      "loss": 0.3533,
+      "step": 1368
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.838170993418172,
+      "learning_rate": 5.524535843659658e-06,
+      "loss": 0.2996,
+      "step": 1369
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.550231203448934,
+      "learning_rate": 5.518855034608187e-06,
+      "loss": 0.3169,
+      "step": 1370
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.232612303022613,
+      "learning_rate": 5.513173548413789e-06,
+      "loss": 0.2769,
+      "step": 1371
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.446267730881251,
+      "learning_rate": 5.507491392491213e-06,
+      "loss": 0.3508,
+      "step": 1372
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.31067411733204,
+      "learning_rate": 5.5018085742560745e-06,
+      "loss": 0.3357,
+      "step": 1373
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.058854724171599,
+      "learning_rate": 5.496125101124862e-06,
+      "loss": 0.2365,
+      "step": 1374
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.759979524744065,
+      "learning_rate": 5.490440980514913e-06,
+      "loss": 0.312,
+      "step": 1375
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.702061973513584,
+      "learning_rate": 5.484756219844408e-06,
+      "loss": 0.2527,
+      "step": 1376
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 19.79957270732681,
+      "learning_rate": 5.47907082653237e-06,
+      "loss": 0.4062,
+      "step": 1377
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.529105258063318,
+      "learning_rate": 5.473384807998644e-06,
+      "loss": 0.3489,
+      "step": 1378
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.259508210832205,
+      "learning_rate": 5.467698171663889e-06,
+      "loss": 0.3208,
+      "step": 1379
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.32824548802698,
+      "learning_rate": 5.462010924949574e-06,
+      "loss": 0.3071,
+      "step": 1380
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.025122499361139,
+      "learning_rate": 5.4563230752779595e-06,
+      "loss": 0.2922,
+      "step": 1381
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.030052556861055,
+      "learning_rate": 5.450634630072099e-06,
+      "loss": 0.4561,
+      "step": 1382
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.196979297683298,
+      "learning_rate": 5.444945596755819e-06,
+      "loss": 0.3584,
+      "step": 1383
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.657005805851943,
+      "learning_rate": 5.439255982753717e-06,
+      "loss": 0.2893,
+      "step": 1384
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.825644998693484,
+      "learning_rate": 5.433565795491142e-06,
+      "loss": 0.2546,
+      "step": 1385
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.83655016316379,
+      "learning_rate": 5.4278750423942e-06,
+      "loss": 0.2949,
+      "step": 1386
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.340846223186169,
+      "learning_rate": 5.422183730889726e-06,
+      "loss": 0.3376,
+      "step": 1387
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.467936049473673,
+      "learning_rate": 5.416491868405293e-06,
+      "loss": 0.313,
+      "step": 1388
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.926174374154367,
+      "learning_rate": 5.410799462369186e-06,
+      "loss": 0.2422,
+      "step": 1389
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.036234456795924,
+      "learning_rate": 5.405106520210401e-06,
+      "loss": 0.3047,
+      "step": 1390
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.027123613980525,
+      "learning_rate": 5.3994130493586385e-06,
+      "loss": 0.312,
+      "step": 1391
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.14281902032603,
+      "learning_rate": 5.393719057244282e-06,
+      "loss": 0.2751,
+      "step": 1392
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.601370729076127,
+      "learning_rate": 5.388024551298398e-06,
+      "loss": 0.353,
+      "step": 1393
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.813099191716484,
+      "learning_rate": 5.382329538952728e-06,
+      "loss": 0.3562,
+      "step": 1394
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.430256347795402,
+      "learning_rate": 5.376634027639664e-06,
+      "loss": 0.2644,
+      "step": 1395
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.366507862686841,
+      "learning_rate": 5.370938024792262e-06,
+      "loss": 0.3054,
+      "step": 1396
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.027369168786956,
+      "learning_rate": 5.365241537844208e-06,
+      "loss": 0.3037,
+      "step": 1397
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.637574243503245,
+      "learning_rate": 5.359544574229829e-06,
+      "loss": 0.3015,
+      "step": 1398
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.744158751301949,
+      "learning_rate": 5.353847141384067e-06,
+      "loss": 0.2607,
+      "step": 1399
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 6.796485449981509,
+      "learning_rate": 5.34814924674248e-06,
+      "loss": 0.3423,
+      "step": 1400
+    },
+    {
+      "epoch": 1.97,
+      "eval_avg_AUC": 0.7637015158995223,
+      "eval_avg_Accuracy": 0.6528514588859416,
+      "eval_avg_Accuracy-right": 0.9649797834876744,
+      "eval_avg_Accuracy-wrong": 0.1085967705253582,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6437556588682916,
+      "eval_last_AUC": 0.7818083403083886,
+      "eval_last_Accuracy": 0.7175066312997348,
+      "eval_last_Accuracy-right": 0.7974435894091562,
+      "eval_last_Accuracy-wrong": 0.578121446440755,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6462244339648467,
+      "eval_max_AUC": 0.6021513751016372,
+      "eval_max_Accuracy": 0.6350712864721485,
+      "eval_max_Accuracy-right": 0.9946524064171123,
+      "eval_max_Accuracy-wrong": 0.008073686604503071,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5359891325751334,
+      "eval_min_AUC": 0.7802640627395767,
+      "eval_min_Accuracy": 0.717092175066313,
+      "eval_min_Accuracy-right": 0.7901395591496022,
+      "eval_min_Accuracy-wrong": 0.5897202638162383,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6474482537920089,
+      "eval_prod_AUC": 0.7762835053685507,
+      "eval_prod_Accuracy": 0.679915450928382,
+      "eval_prod_Accuracy-right": 0.6083865918873093,
+      "eval_prod_Accuracy-wrong": 0.8046395269501934,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6456820655145507,
+      "eval_runtime": 668.4416,
+      "eval_samples_per_second": 36.096,
+      "eval_steps_per_second": 1.128,
+      "eval_sum_AUC": 0.5790593598994821,
+      "eval_sum_Accuracy": 0.6348226127320955,
+      "eval_sum_Accuracy-right": 0.9970653449850007,
+      "eval_sum_Accuracy-wrong": 0.0031839890834659997,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.614199192825227,
+      "step": 1400
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.356084401326975,
+      "learning_rate": 5.3424508977412285e-06,
+      "loss": 0.2542,
+      "step": 1401
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.317410787192379,
+      "learning_rate": 5.336752101817067e-06,
+      "loss": 0.2156,
+      "step": 1402
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.367733745318302,
+      "learning_rate": 5.331052866407328e-06,
+      "loss": 0.2644,
+      "step": 1403
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.7785420469474715,
+      "learning_rate": 5.325353198949924e-06,
+      "loss": 0.2063,
+      "step": 1404
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.158775892415646,
+      "learning_rate": 5.319653106883328e-06,
+      "loss": 0.3301,
+      "step": 1405
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 17.66745649909212,
+      "learning_rate": 5.3139525976465675e-06,
+      "loss": 0.4075,
+      "step": 1406
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.256666043827668,
+      "learning_rate": 5.3082516786792144e-06,
+      "loss": 0.2632,
+      "step": 1407
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.421605478333015,
+      "learning_rate": 5.302550357421378e-06,
+      "loss": 0.2944,
+      "step": 1408
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.24324787985975,
+      "learning_rate": 5.296848641313688e-06,
+      "loss": 0.3098,
+      "step": 1409
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.282736165798594,
+      "learning_rate": 5.291146537797291e-06,
+      "loss": 0.3076,
+      "step": 1410
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.755581840132637,
+      "learning_rate": 5.285444054313841e-06,
+      "loss": 0.2705,
+      "step": 1411
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 14.175281947437268,
+      "learning_rate": 5.279741198305488e-06,
+      "loss": 0.3081,
+      "step": 1412
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.130294340751736,
+      "learning_rate": 5.274037977214864e-06,
+      "loss": 0.3201,
+      "step": 1413
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.580123701581341,
+      "learning_rate": 5.268334398485082e-06,
+      "loss": 0.3591,
+      "step": 1414
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.081218476609858,
+      "learning_rate": 5.262630469559722e-06,
+      "loss": 0.3315,
+      "step": 1415
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.857300981734169,
+      "learning_rate": 5.2569261978828155e-06,
+      "loss": 0.3682,
+      "step": 1416
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.8111412876735224,
+      "learning_rate": 5.251221590898848e-06,
+      "loss": 0.3184,
+      "step": 1417
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.340430051427998,
+      "learning_rate": 5.245516656052738e-06,
+      "loss": 0.2705,
+      "step": 1418
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.47394279903575,
+      "learning_rate": 5.2398114007898346e-06,
+      "loss": 0.1893,
+      "step": 1419
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.926816063998657,
+      "learning_rate": 5.234105832555904e-06,
+      "loss": 0.2166,
+      "step": 1420
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.456135911266939,
+      "learning_rate": 5.228399958797117e-06,
+      "loss": 0.2034,
+      "step": 1421
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.931734391344389,
+      "learning_rate": 5.222693786960053e-06,
+      "loss": 0.2209,
+      "step": 1422
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.379063302414142,
+      "learning_rate": 5.216987324491669e-06,
+      "loss": 0.12,
+      "step": 1423
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.865097696969167,
+      "learning_rate": 5.21128057883931e-06,
+      "loss": 0.1899,
+      "step": 1424
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 7.072822022465626,
+      "learning_rate": 5.205573557450685e-06,
+      "loss": 0.2271,
+      "step": 1425
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 9.768994983712762,
+      "learning_rate": 5.199866267773868e-06,
+      "loss": 0.2751,
+      "step": 1426
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.116652339944973,
+      "learning_rate": 5.194158717257276e-06,
+      "loss": 0.2058,
+      "step": 1427
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.114684331062182,
+      "learning_rate": 5.188450913349674e-06,
+      "loss": 0.155,
+      "step": 1428
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 6.074389965015014,
+      "learning_rate": 5.182742863500152e-06,
+      "loss": 0.2131,
+      "step": 1429
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.6988733037795765,
+      "learning_rate": 5.177034575158125e-06,
+      "loss": 0.1503,
+      "step": 1430
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 6.961739352312222,
+      "learning_rate": 5.171326055773318e-06,
+      "loss": 0.181,
+      "step": 1431
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.202089776152064,
+      "learning_rate": 5.165617312795755e-06,
+      "loss": 0.188,
+      "step": 1432
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.454053459187849,
+      "learning_rate": 5.159908353675754e-06,
+      "loss": 0.1721,
+      "step": 1433
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 11.145524614367577,
+      "learning_rate": 5.154199185863917e-06,
+      "loss": 0.1809,
+      "step": 1434
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.117582797855622,
+      "learning_rate": 5.148489816811114e-06,
+      "loss": 0.1785,
+      "step": 1435
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 14.118056366581786,
+      "learning_rate": 5.142780253968481e-06,
+      "loss": 0.1296,
+      "step": 1436
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.495555840157433,
+      "learning_rate": 5.137070504787407e-06,
+      "loss": 0.1345,
+      "step": 1437
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 9.996979218721604,
+      "learning_rate": 5.1313605767195204e-06,
+      "loss": 0.1941,
+      "step": 1438
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 10.250629116904918,
+      "learning_rate": 5.1256504772166885e-06,
+      "loss": 0.1816,
+      "step": 1439
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 9.554297626953067,
+      "learning_rate": 5.119940213730997e-06,
+      "loss": 0.1616,
+      "step": 1440
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 11.177451585379496,
+      "learning_rate": 5.114229793714749e-06,
+      "loss": 0.1693,
+      "step": 1441
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 11.612392080064051,
+      "learning_rate": 5.1085192246204505e-06,
+      "loss": 0.1849,
+      "step": 1442
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.362700206874424,
+      "learning_rate": 5.102808513900805e-06,
+      "loss": 0.1777,
+      "step": 1443
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.035062883182087,
+      "learning_rate": 5.0970976690086946e-06,
+      "loss": 0.1753,
+      "step": 1444
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.914058672966101,
+      "learning_rate": 5.0913866973971835e-06,
+      "loss": 0.1437,
+      "step": 1445
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.703484973619518,
+      "learning_rate": 5.085675606519496e-06,
+      "loss": 0.149,
+      "step": 1446
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.51875239342496,
+      "learning_rate": 5.079964403829016e-06,
+      "loss": 0.167,
+      "step": 1447
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 7.682166506209115,
+      "learning_rate": 5.07425309677927e-06,
+      "loss": 0.1405,
+      "step": 1448
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 8.112876457695762,
+      "learning_rate": 5.068541692823923e-06,
+      "loss": 0.1459,
+      "step": 1449
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.064008389774031,
+      "learning_rate": 5.062830199416764e-06,
+      "loss": 0.2278,
+      "step": 1450
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.405081945117391,
+      "learning_rate": 5.057118624011702e-06,
+      "loss": 0.2224,
+      "step": 1451
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 10.589590627615484,
+      "learning_rate": 5.051406974062751e-06,
+      "loss": 0.1711,
+      "step": 1452
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 11.16877901244086,
+      "learning_rate": 5.045695257024022e-06,
+      "loss": 0.1656,
+      "step": 1453
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.955824232451285,
+      "learning_rate": 5.039983480349715e-06,
+      "loss": 0.1516,
+      "step": 1454
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 9.824069581591502,
+      "learning_rate": 5.034271651494107e-06,
+      "loss": 0.1448,
+      "step": 1455
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.125173058222917,
+      "learning_rate": 5.028559777911543e-06,
+      "loss": 0.0914,
+      "step": 1456
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 11.067885044638848,
+      "learning_rate": 5.022847867056425e-06,
+      "loss": 0.1654,
+      "step": 1457
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.594275138366642,
+      "learning_rate": 5.017135926383209e-06,
+      "loss": 0.193,
+      "step": 1458
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 11.829333272302115,
+      "learning_rate": 5.011423963346382e-06,
+      "loss": 0.1787,
+      "step": 1459
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 12.128012701282183,
+      "learning_rate": 5.005711985400465e-06,
+      "loss": 0.1743,
+      "step": 1460
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 10.850697799321702,
+      "learning_rate": 5e-06,
+      "loss": 0.1731,
+      "step": 1461
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.404347551133602,
+      "learning_rate": 4.9942880145995355e-06,
+      "loss": 0.1379,
+      "step": 1462
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.885358675916006,
+      "learning_rate": 4.9885760366536205e-06,
+      "loss": 0.1395,
+      "step": 1463
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.4449143212750775,
+      "learning_rate": 4.982864073616792e-06,
+      "loss": 0.124,
+      "step": 1464
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.450214945272208,
+      "learning_rate": 4.977152132943575e-06,
+      "loss": 0.1537,
+      "step": 1465
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.652690167307169,
+      "learning_rate": 4.971440222088459e-06,
+      "loss": 0.1752,
+      "step": 1466
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.283564962601414,
+      "learning_rate": 4.965728348505895e-06,
+      "loss": 0.09,
+      "step": 1467
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 9.681933489739365,
+      "learning_rate": 4.960016519650287e-06,
+      "loss": 0.1892,
+      "step": 1468
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 11.584576311472986,
+      "learning_rate": 4.954304742975979e-06,
+      "loss": 0.2062,
+      "step": 1469
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.427738452789225,
+      "learning_rate": 4.94859302593725e-06,
+      "loss": 0.1482,
+      "step": 1470
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.797815832616965,
+      "learning_rate": 4.9428813759883e-06,
+      "loss": 0.1241,
+      "step": 1471
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.863490393706728,
+      "learning_rate": 4.937169800583237e-06,
+      "loss": 0.139,
+      "step": 1472
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.686312691764378,
+      "learning_rate": 4.931458307176079e-06,
+      "loss": 0.1979,
+      "step": 1473
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.208118068915436,
+      "learning_rate": 4.925746903220732e-06,
+      "loss": 0.1567,
+      "step": 1474
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 10.39960910999579,
+      "learning_rate": 4.920035596170984e-06,
+      "loss": 0.187,
+      "step": 1475
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.613466302771496,
+      "learning_rate": 4.914324393480504e-06,
+      "loss": 0.1592,
+      "step": 1476
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 15.0284418255482,
+      "learning_rate": 4.908613302602817e-06,
+      "loss": 0.231,
+      "step": 1477
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.624605231298514,
+      "learning_rate": 4.902902330991306e-06,
+      "loss": 0.1327,
+      "step": 1478
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 12.510446013495347,
+      "learning_rate": 4.897191486099198e-06,
+      "loss": 0.2125,
+      "step": 1479
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.84140234085107,
+      "learning_rate": 4.891480775379551e-06,
+      "loss": 0.1547,
+      "step": 1480
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.972204992215834,
+      "learning_rate": 4.8857702062852515e-06,
+      "loss": 0.1333,
+      "step": 1481
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.757199003308468,
+      "learning_rate": 4.880059786269005e-06,
+      "loss": 0.1648,
+      "step": 1482
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.613182479998015,
+      "learning_rate": 4.874349522783313e-06,
+      "loss": 0.1895,
+      "step": 1483
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.599088504427954,
+      "learning_rate": 4.86863942328048e-06,
+      "loss": 0.1667,
+      "step": 1484
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.462609318412381,
+      "learning_rate": 4.8629294952125955e-06,
+      "loss": 0.2708,
+      "step": 1485
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.664709934788758,
+      "learning_rate": 4.85721974603152e-06,
+      "loss": 0.1217,
+      "step": 1486
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 9.787261523953665,
+      "learning_rate": 4.851510183188887e-06,
+      "loss": 0.1481,
+      "step": 1487
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 6.565233339024631,
+      "learning_rate": 4.845800814136084e-06,
+      "loss": 0.0999,
+      "step": 1488
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 16.669791665760528,
+      "learning_rate": 4.840091646324248e-06,
+      "loss": 0.2262,
+      "step": 1489
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 19.756592790576974,
+      "learning_rate": 4.834382687204246e-06,
+      "loss": 0.2542,
+      "step": 1490
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 6.285473773707108,
+      "learning_rate": 4.828673944226684e-06,
+      "loss": 0.1234,
+      "step": 1491
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 6.098029852247553,
+      "learning_rate": 4.8229654248418755e-06,
+      "loss": 0.1306,
+      "step": 1492
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 10.331525808813204,
+      "learning_rate": 4.8172571364998485e-06,
+      "loss": 0.183,
+      "step": 1493
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.189281934962404,
+      "learning_rate": 4.811549086650327e-06,
+      "loss": 0.1443,
+      "step": 1494
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 8.970895765050788,
+      "learning_rate": 4.805841282742726e-06,
+      "loss": 0.1589,
+      "step": 1495
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.38118267336291,
+      "learning_rate": 4.800133732226135e-06,
+      "loss": 0.1677,
+      "step": 1496
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.587780855401675,
+      "learning_rate": 4.794426442549317e-06,
+      "loss": 0.1974,
+      "step": 1497
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 8.097658205220087,
+      "learning_rate": 4.788719421160691e-06,
+      "loss": 0.1633,
+      "step": 1498
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 12.099338144780587,
+      "learning_rate": 4.7830126755083325e-06,
+      "loss": 0.1843,
+      "step": 1499
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 9.808076346936454,
+      "learning_rate": 4.77730621303995e-06,
+      "loss": 0.1525,
+      "step": 1500
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.285750449275628,
+      "learning_rate": 4.771600041202884e-06,
+      "loss": 0.1694,
+      "step": 1501
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.826726130071663,
+      "learning_rate": 4.7658941674440995e-06,
+      "loss": 0.1885,
+      "step": 1502
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 12.244722228059112,
+      "learning_rate": 4.760188599210166e-06,
+      "loss": 0.1866,
+      "step": 1503
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.689434064530454,
+      "learning_rate": 4.754483343947262e-06,
+      "loss": 0.093,
+      "step": 1504
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.597312783351038,
+      "learning_rate": 4.748778409101153e-06,
+      "loss": 0.1138,
+      "step": 1505
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 12.302277730154959,
+      "learning_rate": 4.743073802117185e-06,
+      "loss": 0.15,
+      "step": 1506
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 15.614947218910785,
+      "learning_rate": 4.737369530440279e-06,
+      "loss": 0.1931,
+      "step": 1507
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.686860001736905,
+      "learning_rate": 4.731665601514919e-06,
+      "loss": 0.1653,
+      "step": 1508
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 9.224254969452991,
+      "learning_rate": 4.7259620227851375e-06,
+      "loss": 0.1946,
+      "step": 1509
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 7.07294713929005,
+      "learning_rate": 4.720258801694514e-06,
+      "loss": 0.1295,
+      "step": 1510
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 12.79943624643193,
+      "learning_rate": 4.71455594568616e-06,
+      "loss": 0.1792,
+      "step": 1511
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 9.886118031148161,
+      "learning_rate": 4.708853462202711e-06,
+      "loss": 0.1609,
+      "step": 1512
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 13.604085582735456,
+      "learning_rate": 4.7031513586863145e-06,
+      "loss": 0.166,
+      "step": 1513
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.494803860955258,
+      "learning_rate": 4.697449642578625e-06,
+      "loss": 0.1067,
+      "step": 1514
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 8.063527691172082,
+      "learning_rate": 4.6917483213207855e-06,
+      "loss": 0.1311,
+      "step": 1515
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.032980502627158,
+      "learning_rate": 4.686047402353433e-06,
+      "loss": 0.1835,
+      "step": 1516
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.212701715150473,
+      "learning_rate": 4.680346893116673e-06,
+      "loss": 0.1531,
+      "step": 1517
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 13.48686946365647,
+      "learning_rate": 4.674646801050078e-06,
+      "loss": 0.1296,
+      "step": 1518
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.235146686647449,
+      "learning_rate": 4.668947133592674e-06,
+      "loss": 0.188,
+      "step": 1519
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.964258613636722,
+      "learning_rate": 4.663247898182934e-06,
+      "loss": 0.1041,
+      "step": 1520
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.330719303073508,
+      "learning_rate": 4.6575491022587714e-06,
+      "loss": 0.1446,
+      "step": 1521
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.225917603400275,
+      "learning_rate": 4.651850753257521e-06,
+      "loss": 0.1765,
+      "step": 1522
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.62251057695449,
+      "learning_rate": 4.646152858615935e-06,
+      "loss": 0.1726,
+      "step": 1523
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 9.23066370513094,
+      "learning_rate": 4.640455425770174e-06,
+      "loss": 0.171,
+      "step": 1524
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.890504242635467,
+      "learning_rate": 4.6347584621557935e-06,
+      "loss": 0.1836,
+      "step": 1525
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 7.261913989749687,
+      "learning_rate": 4.62906197520774e-06,
+      "loss": 0.1191,
+      "step": 1526
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.119785341911312,
+      "learning_rate": 4.6233659723603374e-06,
+      "loss": 0.1895,
+      "step": 1527
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 7.481693337049464,
+      "learning_rate": 4.617670461047275e-06,
+      "loss": 0.127,
+      "step": 1528
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 8.843023504163401,
+      "learning_rate": 4.611975448701602e-06,
+      "loss": 0.1115,
+      "step": 1529
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 10.04824706426491,
+      "learning_rate": 4.60628094275572e-06,
+      "loss": 0.1552,
+      "step": 1530
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 8.978454584652212,
+      "learning_rate": 4.600586950641362e-06,
+      "loss": 0.1576,
+      "step": 1531
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 10.806075412522567,
+      "learning_rate": 4.594893479789599e-06,
+      "loss": 0.1594,
+      "step": 1532
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.621008483644305,
+      "learning_rate": 4.589200537630815e-06,
+      "loss": 0.2004,
+      "step": 1533
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 9.583296614162675,
+      "learning_rate": 4.583508131594708e-06,
+      "loss": 0.1497,
+      "step": 1534
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 7.641513836194793,
+      "learning_rate": 4.577816269110275e-06,
+      "loss": 0.1248,
+      "step": 1535
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.657473477650932,
+      "learning_rate": 4.572124957605803e-06,
+      "loss": 0.0952,
+      "step": 1536
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.208791120443285,
+      "learning_rate": 4.566434204508858e-06,
+      "loss": 0.1658,
+      "step": 1537
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.96955037761446,
+      "learning_rate": 4.560744017246284e-06,
+      "loss": 0.1389,
+      "step": 1538
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 9.539456538038802,
+      "learning_rate": 4.5550544032441815e-06,
+      "loss": 0.1453,
+      "step": 1539
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 14.670222159296872,
+      "learning_rate": 4.549365369927903e-06,
+      "loss": 0.14,
+      "step": 1540
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 11.309265041516575,
+      "learning_rate": 4.543676924722042e-06,
+      "loss": 0.1698,
+      "step": 1541
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 9.509405850807367,
+      "learning_rate": 4.537989075050429e-06,
+      "loss": 0.1591,
+      "step": 1542
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 12.854004441651064,
+      "learning_rate": 4.532301828336112e-06,
+      "loss": 0.1821,
+      "step": 1543
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.96901820461623,
+      "learning_rate": 4.526615192001357e-06,
+      "loss": 0.1328,
+      "step": 1544
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 11.368130027712281,
+      "learning_rate": 4.520929173467632e-06,
+      "loss": 0.2025,
+      "step": 1545
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 16.001037688387765,
+      "learning_rate": 4.515243780155594e-06,
+      "loss": 0.167,
+      "step": 1546
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 10.249101871809634,
+      "learning_rate": 4.50955901948509e-06,
+      "loss": 0.1609,
+      "step": 1547
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 8.434891580809932,
+      "learning_rate": 4.503874898875139e-06,
+      "loss": 0.1141,
+      "step": 1548
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 12.18526222848323,
+      "learning_rate": 4.4981914257439254e-06,
+      "loss": 0.1475,
+      "step": 1549
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 8.548140174826209,
+      "learning_rate": 4.492508607508788e-06,
+      "loss": 0.1841,
+      "step": 1550
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 9.792516797023582,
+      "learning_rate": 4.4868264515862115e-06,
+      "loss": 0.1224,
+      "step": 1551
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 10.357885916011186,
+      "learning_rate": 4.481144965391815e-06,
+      "loss": 0.1213,
+      "step": 1552
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.297797975646041,
+      "learning_rate": 4.4754641563403435e-06,
+      "loss": 0.1923,
+      "step": 1553
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 7.241327077463006,
+      "learning_rate": 4.469784031845658e-06,
+      "loss": 0.1478,
+      "step": 1554
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 22.816223603210688,
+      "learning_rate": 4.464104599320732e-06,
+      "loss": 0.2216,
+      "step": 1555
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.859611416813832,
+      "learning_rate": 4.458425866177628e-06,
+      "loss": 0.1503,
+      "step": 1556
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.335839939799229,
+      "learning_rate": 4.452747839827498e-06,
+      "loss": 0.167,
+      "step": 1557
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 11.455929618268724,
+      "learning_rate": 4.4470705276805745e-06,
+      "loss": 0.1754,
+      "step": 1558
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.08070178651429,
+      "learning_rate": 4.441393937146152e-06,
+      "loss": 0.1707,
+      "step": 1559
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 9.570023615186713,
+      "learning_rate": 4.4357180756325915e-06,
+      "loss": 0.1455,
+      "step": 1560
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.433810793007616,
+      "learning_rate": 4.430042950547298e-06,
+      "loss": 0.1863,
+      "step": 1561
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 13.253389807632356,
+      "learning_rate": 4.424368569296713e-06,
+      "loss": 0.1838,
+      "step": 1562
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 13.04982618905267,
+      "learning_rate": 4.418694939286311e-06,
+      "loss": 0.2079,
+      "step": 1563
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 9.176554982044209,
+      "learning_rate": 4.413022067920585e-06,
+      "loss": 0.1815,
+      "step": 1564
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 8.367187033024646,
+      "learning_rate": 4.4073499626030365e-06,
+      "loss": 0.1599,
+      "step": 1565
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 8.61041069617733,
+      "learning_rate": 4.401678630736172e-06,
+      "loss": 0.1499,
+      "step": 1566
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 7.481682738045144,
+      "learning_rate": 4.396008079721481e-06,
+      "loss": 0.1216,
+      "step": 1567
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 8.380280328091251,
+      "learning_rate": 4.390338316959442e-06,
+      "loss": 0.1296,
+      "step": 1568
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 8.333075943931089,
+      "learning_rate": 4.384669349849498e-06,
+      "loss": 0.1648,
+      "step": 1569
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.804263834997744,
+      "learning_rate": 4.379001185790056e-06,
+      "loss": 0.1858,
+      "step": 1570
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.0741280597781575,
+      "learning_rate": 4.373333832178478e-06,
+      "loss": 0.1243,
+      "step": 1571
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 13.863650705471029,
+      "learning_rate": 4.3676672964110664e-06,
+      "loss": 0.188,
+      "step": 1572
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 9.651352870891408,
+      "learning_rate": 4.362001585883053e-06,
+      "loss": 0.2136,
+      "step": 1573
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 9.928419104997293,
+      "learning_rate": 4.356336707988595e-06,
+      "loss": 0.1715,
+      "step": 1574
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 10.54742443566365,
+      "learning_rate": 4.350672670120766e-06,
+      "loss": 0.167,
+      "step": 1575
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 13.200358790762921,
+      "learning_rate": 4.3450094796715354e-06,
+      "loss": 0.2327,
+      "step": 1576
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 16.482855880669838,
+      "learning_rate": 4.339347144031778e-06,
+      "loss": 0.2245,
+      "step": 1577
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.860664803303719,
+      "learning_rate": 4.333685670591244e-06,
+      "loss": 0.1689,
+      "step": 1578
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 7.718170573450203,
+      "learning_rate": 4.32802506673856e-06,
+      "loss": 0.1514,
+      "step": 1579
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.08099983030036,
+      "learning_rate": 4.322365339861221e-06,
+      "loss": 0.2041,
+      "step": 1580
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 10.004311535176488,
+      "learning_rate": 4.316706497345572e-06,
+      "loss": 0.1697,
+      "step": 1581
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.079151301326064,
+      "learning_rate": 4.31104854657681e-06,
+      "loss": 0.2095,
+      "step": 1582
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.29250871911384,
+      "learning_rate": 4.305391494938965e-06,
+      "loss": 0.2142,
+      "step": 1583
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 10.95110385339461,
+      "learning_rate": 4.299735349814894e-06,
+      "loss": 0.1832,
+      "step": 1584
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.828655336783058,
+      "learning_rate": 4.29408011858627e-06,
+      "loss": 0.1248,
+      "step": 1585
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.47581810237909,
+      "learning_rate": 4.2884258086335755e-06,
+      "loss": 0.1385,
+      "step": 1586
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 11.10254942422425,
+      "learning_rate": 4.282772427336087e-06,
+      "loss": 0.161,
+      "step": 1587
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 11.939410901025518,
+      "learning_rate": 4.277119982071873e-06,
+      "loss": 0.1525,
+      "step": 1588
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.45308384838439,
+      "learning_rate": 4.27146848021778e-06,
+      "loss": 0.1328,
+      "step": 1589
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 12.291295376000788,
+      "learning_rate": 4.265817929149421e-06,
+      "loss": 0.2527,
+      "step": 1590
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.185104093908278,
+      "learning_rate": 4.260168336241169e-06,
+      "loss": 0.1346,
+      "step": 1591
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 9.721611633014788,
+      "learning_rate": 4.254519708866147e-06,
+      "loss": 0.1006,
+      "step": 1592
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.545000987162355,
+      "learning_rate": 4.248872054396215e-06,
+      "loss": 0.2476,
+      "step": 1593
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.522893948444759,
+      "learning_rate": 4.24322538020197e-06,
+      "loss": 0.1615,
+      "step": 1594
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 13.467993788986108,
+      "learning_rate": 4.237579693652725e-06,
+      "loss": 0.1926,
+      "step": 1595
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.539204986333665,
+      "learning_rate": 4.231935002116504e-06,
+      "loss": 0.2255,
+      "step": 1596
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 12.856800308790845,
+      "learning_rate": 4.2262913129600326e-06,
+      "loss": 0.1809,
+      "step": 1597
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.72501517676801,
+      "learning_rate": 4.220648633548726e-06,
+      "loss": 0.1271,
+      "step": 1598
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 7.465505619139734,
+      "learning_rate": 4.215006971246691e-06,
+      "loss": 0.1366,
+      "step": 1599
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 7.20035194577559,
+      "learning_rate": 4.209366333416697e-06,
+      "loss": 0.0953,
+      "step": 1600
+    },
+    {
+      "epoch": 2.26,
+      "eval_avg_AUC": 0.7548414651499317,
+      "eval_avg_Accuracy": 0.6567059018567639,
+      "eval_avg_Accuracy-right": 0.9458719186122343,
+      "eval_avg_Accuracy-wrong": 0.15249033431885375,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6521147410430262,
+      "eval_last_AUC": 0.7806854990147515,
+      "eval_last_Accuracy": 0.7216097480106101,
+      "eval_last_Accuracy-right": 0.7957480109560454,
+      "eval_last_Accuracy-wrong": 0.592335683420514,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6639303542682262,
+      "eval_max_AUC": 0.5900950509938672,
+      "eval_max_Accuracy": 0.6347397214854111,
+      "eval_max_Accuracy-right": 0.9900221729490022,
+      "eval_max_Accuracy-wrong": 0.01523766204230157,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5495290106770748,
+      "eval_min_AUC": 0.7752218110638355,
+      "eval_min_Accuracy": 0.7151442307692307,
+      "eval_min_Accuracy-right": 0.7752054258510499,
+      "eval_min_Accuracy-wrong": 0.6104161928587674,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6608959915407909,
+      "eval_prod_AUC": 0.7720012732040791,
+      "eval_prod_Accuracy": 0.6965351458885941,
+      "eval_prod_Accuracy-right": 0.6742532933350723,
+      "eval_prod_Accuracy-wrong": 0.7353877643848078,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6562137495808593,
+      "eval_runtime": 668.1261,
+      "eval_samples_per_second": 36.113,
+      "eval_steps_per_second": 1.129,
+      "eval_sum_AUC": 0.5765936653217468,
+      "eval_sum_Accuracy": 0.6353199602122016,
+      "eval_sum_Accuracy-right": 0.9945219773053345,
+      "eval_sum_Accuracy-wrong": 0.008983397771207642,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6163810336148485,
+      "step": 1600
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 10.914778664351802,
+      "learning_rate": 4.2037267274201785e-06,
+      "loss": 0.1733,
+      "step": 1601
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 7.174218144600878,
+      "learning_rate": 4.198088160617225e-06,
+      "loss": 0.1206,
+      "step": 1602
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 6.6770386901760865,
+      "learning_rate": 4.192450640366572e-06,
+      "loss": 0.1284,
+      "step": 1603
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 8.688209616989877,
+      "learning_rate": 4.186814174025582e-06,
+      "loss": 0.1729,
+      "step": 1604
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 10.12353275757718,
+      "learning_rate": 4.181178768950251e-06,
+      "loss": 0.1735,
+      "step": 1605
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 10.536050534420017,
+      "learning_rate": 4.175544432495184e-06,
+      "loss": 0.2021,
+      "step": 1606
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 9.339188919970905,
+      "learning_rate": 4.169911172013594e-06,
+      "loss": 0.1805,
+      "step": 1607
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 8.192075473407726,
+      "learning_rate": 4.164278994857287e-06,
+      "loss": 0.1134,
+      "step": 1608
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 9.89493972794337,
+      "learning_rate": 4.158647908376659e-06,
+      "loss": 0.1843,
+      "step": 1609
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 11.306073883513822,
+      "learning_rate": 4.153017919920676e-06,
+      "loss": 0.2013,
+      "step": 1610
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 8.23943833468193,
+      "learning_rate": 4.147389036836881e-06,
+      "loss": 0.118,
+      "step": 1611
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 9.074950902818875,
+      "learning_rate": 4.141761266471368e-06,
+      "loss": 0.1487,
+      "step": 1612
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 10.309640494637124,
+      "learning_rate": 4.1361346161687775e-06,
+      "loss": 0.179,
+      "step": 1613
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 8.283605416226944,
+      "learning_rate": 4.130509093272293e-06,
+      "loss": 0.1733,
+      "step": 1614
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 7.477546638088935,
+      "learning_rate": 4.124884705123619e-06,
+      "loss": 0.1145,
+      "step": 1615
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 8.417713818835988,
+      "learning_rate": 4.119261459062992e-06,
+      "loss": 0.1644,
+      "step": 1616
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 9.663192272637152,
+      "learning_rate": 4.113639362429146e-06,
+      "loss": 0.1414,
+      "step": 1617
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 8.353641051363024,
+      "learning_rate": 4.108018422559319e-06,
+      "loss": 0.0979,
+      "step": 1618
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.081802200299624,
+      "learning_rate": 4.102398646789239e-06,
+      "loss": 0.1992,
+      "step": 1619
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 8.790331366575126,
+      "learning_rate": 4.096780042453118e-06,
+      "loss": 0.1483,
+      "step": 1620
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 7.076051862597402,
+      "learning_rate": 4.091162616883634e-06,
+      "loss": 0.1234,
+      "step": 1621
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.665349724559903,
+      "learning_rate": 4.085546377411931e-06,
+      "loss": 0.1348,
+      "step": 1622
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 11.855307437445534,
+      "learning_rate": 4.0799313313676e-06,
+      "loss": 0.1962,
+      "step": 1623
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 10.656473967181435,
+      "learning_rate": 4.074317486078682e-06,
+      "loss": 0.1514,
+      "step": 1624
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 7.830658209856807,
+      "learning_rate": 4.068704848871644e-06,
+      "loss": 0.1454,
+      "step": 1625
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 17.69106836478689,
+      "learning_rate": 4.063093427071376e-06,
+      "loss": 0.2837,
+      "step": 1626
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 10.441323522726666,
+      "learning_rate": 4.057483228001191e-06,
+      "loss": 0.1528,
+      "step": 1627
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.216865822931505,
+      "learning_rate": 4.051874258982797e-06,
+      "loss": 0.1598,
+      "step": 1628
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 9.244596937281933,
+      "learning_rate": 4.0462665273363e-06,
+      "loss": 0.1606,
+      "step": 1629
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.778039054850359,
+      "learning_rate": 4.04066004038019e-06,
+      "loss": 0.1625,
+      "step": 1630
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 9.985898182762222,
+      "learning_rate": 4.0350548054313346e-06,
+      "loss": 0.1458,
+      "step": 1631
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.278422282316042,
+      "learning_rate": 4.029450829804962e-06,
+      "loss": 0.134,
+      "step": 1632
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.281097292125441,
+      "learning_rate": 4.023848120814668e-06,
+      "loss": 0.1627,
+      "step": 1633
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.452468131013758,
+      "learning_rate": 4.018246685772385e-06,
+      "loss": 0.1559,
+      "step": 1634
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 14.782239586048684,
+      "learning_rate": 4.0126465319883865e-06,
+      "loss": 0.2159,
+      "step": 1635
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.609619527991065,
+      "learning_rate": 4.007047666771274e-06,
+      "loss": 0.1782,
+      "step": 1636
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.084175424518593,
+      "learning_rate": 4.001450097427965e-06,
+      "loss": 0.1836,
+      "step": 1637
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 14.775065099794938,
+      "learning_rate": 3.995853831263691e-06,
+      "loss": 0.2054,
+      "step": 1638
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.093108085951428,
+      "learning_rate": 3.990258875581978e-06,
+      "loss": 0.1611,
+      "step": 1639
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.686306575573022,
+      "learning_rate": 3.9846652376846425e-06,
+      "loss": 0.226,
+      "step": 1640
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 12.05203460315301,
+      "learning_rate": 3.979072924871784e-06,
+      "loss": 0.1658,
+      "step": 1641
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 8.65831867582176,
+      "learning_rate": 3.973481944441771e-06,
+      "loss": 0.14,
+      "step": 1642
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 12.40652383730445,
+      "learning_rate": 3.967892303691228e-06,
+      "loss": 0.1732,
+      "step": 1643
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 8.959194737667682,
+      "learning_rate": 3.962304009915041e-06,
+      "loss": 0.1685,
+      "step": 1644
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.052271739708813,
+      "learning_rate": 3.956717070406331e-06,
+      "loss": 0.167,
+      "step": 1645
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 14.543069374688109,
+      "learning_rate": 3.951131492456455e-06,
+      "loss": 0.1892,
+      "step": 1646
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 12.784162801138347,
+      "learning_rate": 3.945547283354989e-06,
+      "loss": 0.1571,
+      "step": 1647
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.17199307349458,
+      "learning_rate": 3.939964450389728e-06,
+      "loss": 0.1651,
+      "step": 1648
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.706892521894632,
+      "learning_rate": 3.934383000846665e-06,
+      "loss": 0.1624,
+      "step": 1649
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 11.310816473031561,
+      "learning_rate": 3.928802942009995e-06,
+      "loss": 0.1748,
+      "step": 1650
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.162407558635968,
+      "learning_rate": 3.923224281162091e-06,
+      "loss": 0.1515,
+      "step": 1651
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 11.446173241585688,
+      "learning_rate": 3.917647025583505e-06,
+      "loss": 0.2202,
+      "step": 1652
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 13.765701084214786,
+      "learning_rate": 3.912071182552956e-06,
+      "loss": 0.1953,
+      "step": 1653
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.551166777744019,
+      "learning_rate": 3.906496759347313e-06,
+      "loss": 0.1531,
+      "step": 1654
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.14725877485652,
+      "learning_rate": 3.900923763241602e-06,
+      "loss": 0.1475,
+      "step": 1655
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 8.002377153360246,
+      "learning_rate": 3.895352201508981e-06,
+      "loss": 0.1479,
+      "step": 1656
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 7.740583955305286,
+      "learning_rate": 3.889782081420734e-06,
+      "loss": 0.1622,
+      "step": 1657
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 9.78783535548308,
+      "learning_rate": 3.884213410246267e-06,
+      "loss": 0.1534,
+      "step": 1658
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 13.639214942322164,
+      "learning_rate": 3.8786461952530955e-06,
+      "loss": 0.1691,
+      "step": 1659
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 11.805702253865466,
+      "learning_rate": 3.87308044370683e-06,
+      "loss": 0.1808,
+      "step": 1660
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 7.252599759491916,
+      "learning_rate": 3.867516162871177e-06,
+      "loss": 0.1422,
+      "step": 1661
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 9.492746853888862,
+      "learning_rate": 3.861953360007923e-06,
+      "loss": 0.1809,
+      "step": 1662
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 12.331069489001862,
+      "learning_rate": 3.856392042376919e-06,
+      "loss": 0.2476,
+      "step": 1663
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 6.137088757173362,
+      "learning_rate": 3.850832217236085e-06,
+      "loss": 0.0995,
+      "step": 1664
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 7.659003710167093,
+      "learning_rate": 3.84527389184139e-06,
+      "loss": 0.1417,
+      "step": 1665
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 9.87520448968634,
+      "learning_rate": 3.839717073446842e-06,
+      "loss": 0.1855,
+      "step": 1666
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.385953249250007,
+      "learning_rate": 3.834161769304494e-06,
+      "loss": 0.0994,
+      "step": 1667
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 12.909638835765481,
+      "learning_rate": 3.82860798666441e-06,
+      "loss": 0.1793,
+      "step": 1668
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 11.615388900243833,
+      "learning_rate": 3.823055732774675e-06,
+      "loss": 0.1442,
+      "step": 1669
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 9.549785428400128,
+      "learning_rate": 3.817505014881378e-06,
+      "loss": 0.1538,
+      "step": 1670
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 12.79068551526811,
+      "learning_rate": 3.8119558402285994e-06,
+      "loss": 0.1664,
+      "step": 1671
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 9.42588841457632,
+      "learning_rate": 3.8064082160584138e-06,
+      "loss": 0.1594,
+      "step": 1672
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 9.15863130769092,
+      "learning_rate": 3.8008621496108655e-06,
+      "loss": 0.1406,
+      "step": 1673
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 10.508929397005867,
+      "learning_rate": 3.7953176481239675e-06,
+      "loss": 0.1897,
+      "step": 1674
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 11.395921650439938,
+      "learning_rate": 3.7897747188336897e-06,
+      "loss": 0.1782,
+      "step": 1675
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 16.661771945514047,
+      "learning_rate": 3.7842333689739524e-06,
+      "loss": 0.1799,
+      "step": 1676
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 14.407333316915672,
+      "learning_rate": 3.778693605776613e-06,
+      "loss": 0.146,
+      "step": 1677
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 11.381382121045524,
+      "learning_rate": 3.7731554364714605e-06,
+      "loss": 0.2036,
+      "step": 1678
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.117071653789596,
+      "learning_rate": 3.7676188682861993e-06,
+      "loss": 0.145,
+      "step": 1679
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 7.604839096526782,
+      "learning_rate": 3.762083908446449e-06,
+      "loss": 0.0922,
+      "step": 1680
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.893389106007715,
+      "learning_rate": 3.756550564175727e-06,
+      "loss": 0.1722,
+      "step": 1681
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 11.651587220147947,
+      "learning_rate": 3.7510188426954413e-06,
+      "loss": 0.1625,
+      "step": 1682
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.381530352817052,
+      "learning_rate": 3.7454887512248884e-06,
+      "loss": 0.1415,
+      "step": 1683
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 12.977507212115272,
+      "learning_rate": 3.73996029698123e-06,
+      "loss": 0.1394,
+      "step": 1684
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 11.228873177744479,
+      "learning_rate": 3.734433487179495e-06,
+      "loss": 0.1353,
+      "step": 1685
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 11.439488677667615,
+      "learning_rate": 3.7289083290325668e-06,
+      "loss": 0.1365,
+      "step": 1686
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 12.264730473143407,
+      "learning_rate": 3.7233848297511693e-06,
+      "loss": 0.1421,
+      "step": 1687
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 19.633439607460765,
+      "learning_rate": 3.717862996543864e-06,
+      "loss": 0.1738,
+      "step": 1688
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 9.886057670759039,
+      "learning_rate": 3.712342836617041e-06,
+      "loss": 0.1472,
+      "step": 1689
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 10.99484676660753,
+      "learning_rate": 3.7068243571749025e-06,
+      "loss": 0.135,
+      "step": 1690
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 30.817385145179777,
+      "learning_rate": 3.7013075654194586e-06,
+      "loss": 0.2695,
+      "step": 1691
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 16.38773103394755,
+      "learning_rate": 3.695792468550517e-06,
+      "loss": 0.2295,
+      "step": 1692
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 11.080035005159086,
+      "learning_rate": 3.6902790737656715e-06,
+      "loss": 0.1797,
+      "step": 1693
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.640285273861652,
+      "learning_rate": 3.684767388260301e-06,
+      "loss": 0.1621,
+      "step": 1694
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 7.855728699920788,
+      "learning_rate": 3.6792574192275464e-06,
+      "loss": 0.1603,
+      "step": 1695
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 8.403675159369497,
+      "learning_rate": 3.673749173858312e-06,
+      "loss": 0.1416,
+      "step": 1696
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 8.210390730714098,
+      "learning_rate": 3.6682426593412507e-06,
+      "loss": 0.1208,
+      "step": 1697
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.100146866194457,
+      "learning_rate": 3.6627378828627603e-06,
+      "loss": 0.1559,
+      "step": 1698
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 18.297860275096234,
+      "learning_rate": 3.6572348516069645e-06,
+      "loss": 0.2537,
+      "step": 1699
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 9.17569993939474,
+      "learning_rate": 3.6517335727557147e-06,
+      "loss": 0.1705,
+      "step": 1700
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 11.077899018599467,
+      "learning_rate": 3.646234053488574e-06,
+      "loss": 0.1903,
+      "step": 1701
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 8.586537673182832,
+      "learning_rate": 3.6407363009828078e-06,
+      "loss": 0.1432,
+      "step": 1702
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 11.541952496665932,
+      "learning_rate": 3.635240322413375e-06,
+      "loss": 0.1608,
+      "step": 1703
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 10.517444784350534,
+      "learning_rate": 3.629746124952922e-06,
+      "loss": 0.149,
+      "step": 1704
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 12.446653108109421,
+      "learning_rate": 3.624253715771766e-06,
+      "loss": 0.1614,
+      "step": 1705
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.36020866465334,
+      "learning_rate": 3.618763102037899e-06,
+      "loss": 0.1161,
+      "step": 1706
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 6.7880176910350984,
+      "learning_rate": 3.6132742909169626e-06,
+      "loss": 0.1199,
+      "step": 1707
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 10.4852863160582,
+      "learning_rate": 3.607787289572247e-06,
+      "loss": 0.1476,
+      "step": 1708
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 14.27472734487622,
+      "learning_rate": 3.6023021051646824e-06,
+      "loss": 0.2373,
+      "step": 1709
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 6.517264825394254,
+      "learning_rate": 3.596818744852823e-06,
+      "loss": 0.1096,
+      "step": 1710
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 11.212593152757549,
+      "learning_rate": 3.5913372157928515e-06,
+      "loss": 0.156,
+      "step": 1711
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 13.77530779940902,
+      "learning_rate": 3.585857525138554e-06,
+      "loss": 0.2251,
+      "step": 1712
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.10964500881743,
+      "learning_rate": 3.580379680041317e-06,
+      "loss": 0.1271,
+      "step": 1713
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.15097608642252,
+      "learning_rate": 3.5749036876501196e-06,
+      "loss": 0.1748,
+      "step": 1714
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 8.401449777974427,
+      "learning_rate": 3.5694295551115265e-06,
+      "loss": 0.1358,
+      "step": 1715
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 8.744267556843841,
+      "learning_rate": 3.563957289569669e-06,
+      "loss": 0.1736,
+      "step": 1716
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.668375282649873,
+      "learning_rate": 3.5584868981662463e-06,
+      "loss": 0.1324,
+      "step": 1717
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.520790441127634,
+      "learning_rate": 3.553018388040512e-06,
+      "loss": 0.1675,
+      "step": 1718
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.728068373189192,
+      "learning_rate": 3.54755176632926e-06,
+      "loss": 0.1664,
+      "step": 1719
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.849303166156309,
+      "learning_rate": 3.542087040166825e-06,
+      "loss": 0.1871,
+      "step": 1720
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.560503803655912,
+      "learning_rate": 3.5366242166850624e-06,
+      "loss": 0.1871,
+      "step": 1721
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 14.451365104618764,
+      "learning_rate": 3.5311633030133515e-06,
+      "loss": 0.2119,
+      "step": 1722
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.520628197198828,
+      "learning_rate": 3.5257043062785723e-06,
+      "loss": 0.1454,
+      "step": 1723
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 8.312438316620796,
+      "learning_rate": 3.5202472336051075e-06,
+      "loss": 0.1355,
+      "step": 1724
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.940050159367352,
+      "learning_rate": 3.5147920921148267e-06,
+      "loss": 0.1493,
+      "step": 1725
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 13.194699454250648,
+      "learning_rate": 3.509338888927079e-06,
+      "loss": 0.1576,
+      "step": 1726
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.827420520743791,
+      "learning_rate": 3.5038876311586833e-06,
+      "loss": 0.168,
+      "step": 1727
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.688036246997466,
+      "learning_rate": 3.498438325923924e-06,
+      "loss": 0.1359,
+      "step": 1728
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 14.871825608037007,
+      "learning_rate": 3.492990980334533e-06,
+      "loss": 0.1644,
+      "step": 1729
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 15.816238180084131,
+      "learning_rate": 3.4875456014996857e-06,
+      "loss": 0.1903,
+      "step": 1730
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.485332608103633,
+      "learning_rate": 3.4821021965259905e-06,
+      "loss": 0.1083,
+      "step": 1731
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 10.997038202334883,
+      "learning_rate": 3.4766607725174796e-06,
+      "loss": 0.1404,
+      "step": 1732
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 7.669969730454107,
+      "learning_rate": 3.471221336575601e-06,
+      "loss": 0.1146,
+      "step": 1733
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 11.002486066305389,
+      "learning_rate": 3.465783895799209e-06,
+      "loss": 0.1421,
+      "step": 1734
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 7.901450159587138,
+      "learning_rate": 3.46034845728455e-06,
+      "loss": 0.1108,
+      "step": 1735
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 12.84192598817494,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.1742,
+      "step": 1736
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 11.803669410993303,
+      "learning_rate": 3.4494836154123608e-06,
+      "loss": 0.149,
+      "step": 1737
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.643905423712729,
+      "learning_rate": 3.444054226234222e-06,
+      "loss": 0.162,
+      "step": 1738
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.978533228927965,
+      "learning_rate": 3.4386268676765927e-06,
+      "loss": 0.2097,
+      "step": 1739
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 9.019069030294059,
+      "learning_rate": 3.4332015468225616e-06,
+      "loss": 0.1642,
+      "step": 1740
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 11.299718705374458,
+      "learning_rate": 3.427778270752561e-06,
+      "loss": 0.1458,
+      "step": 1741
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 12.403967767744197,
+      "learning_rate": 3.422357046544353e-06,
+      "loss": 0.2014,
+      "step": 1742
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 12.444006233449489,
+      "learning_rate": 3.4169378812730236e-06,
+      "loss": 0.2135,
+      "step": 1743
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 10.371109404188237,
+      "learning_rate": 3.4115207820109687e-06,
+      "loss": 0.1382,
+      "step": 1744
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.65407336948313,
+      "learning_rate": 3.4061057558278948e-06,
+      "loss": 0.21,
+      "step": 1745
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 9.933035076295866,
+      "learning_rate": 3.400692809790796e-06,
+      "loss": 0.1267,
+      "step": 1746
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 10.341165098601893,
+      "learning_rate": 3.3952819509639534e-06,
+      "loss": 0.1663,
+      "step": 1747
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.270905645597967,
+      "learning_rate": 3.389873186408926e-06,
+      "loss": 0.1514,
+      "step": 1748
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.557284591750292,
+      "learning_rate": 3.384466523184535e-06,
+      "loss": 0.2013,
+      "step": 1749
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 8.970295677858918,
+      "learning_rate": 3.3790619683468673e-06,
+      "loss": 0.137,
+      "step": 1750
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.253160535926262,
+      "learning_rate": 3.3736595289492513e-06,
+      "loss": 0.0967,
+      "step": 1751
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.230402484373842,
+      "learning_rate": 3.3682592120422575e-06,
+      "loss": 0.1292,
+      "step": 1752
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.35820748992435,
+      "learning_rate": 3.362861024673683e-06,
+      "loss": 0.217,
+      "step": 1753
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.895845777707629,
+      "learning_rate": 3.357464973888552e-06,
+      "loss": 0.1016,
+      "step": 1754
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 11.47627452608572,
+      "learning_rate": 3.3520710667290933e-06,
+      "loss": 0.2172,
+      "step": 1755
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.098983310688122,
+      "learning_rate": 3.346679310234744e-06,
+      "loss": 0.1105,
+      "step": 1756
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.931586624355846,
+      "learning_rate": 3.341289711442132e-06,
+      "loss": 0.1172,
+      "step": 1757
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 10.649141036630857,
+      "learning_rate": 3.3359022773850673e-06,
+      "loss": 0.2102,
+      "step": 1758
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 10.067605550424355,
+      "learning_rate": 3.3305170150945387e-06,
+      "loss": 0.1633,
+      "step": 1759
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.232526827915088,
+      "learning_rate": 3.3251339315986965e-06,
+      "loss": 0.1285,
+      "step": 1760
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 11.081878059125591,
+      "learning_rate": 3.319753033922849e-06,
+      "loss": 0.177,
+      "step": 1761
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 9.41535025400267,
+      "learning_rate": 3.3143743290894563e-06,
+      "loss": 0.1565,
+      "step": 1762
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 6.703320684572221,
+      "learning_rate": 3.308997824118111e-06,
+      "loss": 0.1296,
+      "step": 1763
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 9.31139056246412,
+      "learning_rate": 3.3036235260255355e-06,
+      "loss": 0.1814,
+      "step": 1764
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 8.588307858188498,
+      "learning_rate": 3.298251441825574e-06,
+      "loss": 0.1571,
+      "step": 1765
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.788549972944569,
+      "learning_rate": 3.292881578529179e-06,
+      "loss": 0.1763,
+      "step": 1766
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 9.178792684031869,
+      "learning_rate": 3.287513943144409e-06,
+      "loss": 0.1614,
+      "step": 1767
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 8.728388695035205,
+      "learning_rate": 3.282148542676411e-06,
+      "loss": 0.1428,
+      "step": 1768
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 8.78213777109039,
+      "learning_rate": 3.2767853841274154e-06,
+      "loss": 0.149,
+      "step": 1769
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 9.318690471009578,
+      "learning_rate": 3.2714244744967283e-06,
+      "loss": 0.1129,
+      "step": 1770
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 10.093913872528915,
+      "learning_rate": 3.2660658207807205e-06,
+      "loss": 0.171,
+      "step": 1771
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 13.373668231794781,
+      "learning_rate": 3.260709429972816e-06,
+      "loss": 0.1849,
+      "step": 1772
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 8.38231416211088,
+      "learning_rate": 3.2553553090634917e-06,
+      "loss": 0.1129,
+      "step": 1773
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.900386864660603,
+      "learning_rate": 3.2500034650402552e-06,
+      "loss": 0.09,
+      "step": 1774
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 11.389345081920641,
+      "learning_rate": 3.2446539048876474e-06,
+      "loss": 0.1716,
+      "step": 1775
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 11.68510752517661,
+      "learning_rate": 3.2393066355872264e-06,
+      "loss": 0.1593,
+      "step": 1776
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.262413020914167,
+      "learning_rate": 3.233961664117559e-06,
+      "loss": 0.1925,
+      "step": 1777
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 9.761042527451718,
+      "learning_rate": 3.2286189974542194e-06,
+      "loss": 0.137,
+      "step": 1778
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 13.280714864344736,
+      "learning_rate": 3.2232786425697672e-06,
+      "loss": 0.157,
+      "step": 1779
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 9.464608293951072,
+      "learning_rate": 3.217940606433747e-06,
+      "loss": 0.1343,
+      "step": 1780
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 9.780552069096707,
+      "learning_rate": 3.2126048960126787e-06,
+      "loss": 0.1433,
+      "step": 1781
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.427941047472614,
+      "learning_rate": 3.2072715182700464e-06,
+      "loss": 0.1545,
+      "step": 1782
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.567179071797614,
+      "learning_rate": 3.2019404801662858e-06,
+      "loss": 0.1509,
+      "step": 1783
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 7.7593391647862076,
+      "learning_rate": 3.196611788658788e-06,
+      "loss": 0.1343,
+      "step": 1784
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.822857675942496,
+      "learning_rate": 3.191285450701874e-06,
+      "loss": 0.1731,
+      "step": 1785
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.510775942476116,
+      "learning_rate": 3.1859614732467957e-06,
+      "loss": 0.1635,
+      "step": 1786
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.084676092077526,
+      "learning_rate": 3.180639863241726e-06,
+      "loss": 0.1185,
+      "step": 1787
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 11.356637072261833,
+      "learning_rate": 3.1753206276317416e-06,
+      "loss": 0.1567,
+      "step": 1788
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 13.672304594962585,
+      "learning_rate": 3.1700037733588317e-06,
+      "loss": 0.1824,
+      "step": 1789
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.558343908145135,
+      "learning_rate": 3.164689307361869e-06,
+      "loss": 0.1398,
+      "step": 1790
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 14.343488650122312,
+      "learning_rate": 3.1593772365766107e-06,
+      "loss": 0.1836,
+      "step": 1791
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 12.040400681830286,
+      "learning_rate": 3.1540675679356904e-06,
+      "loss": 0.2041,
+      "step": 1792
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 12.357346734325402,
+      "learning_rate": 3.1487603083686036e-06,
+      "loss": 0.2041,
+      "step": 1793
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 11.779499965370906,
+      "learning_rate": 3.1434554648017047e-06,
+      "loss": 0.1781,
+      "step": 1794
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 10.522074950314213,
+      "learning_rate": 3.138153044158194e-06,
+      "loss": 0.1426,
+      "step": 1795
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 13.452320315307569,
+      "learning_rate": 3.1328530533581102e-06,
+      "loss": 0.1953,
+      "step": 1796
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 11.490249686361794,
+      "learning_rate": 3.12755549931832e-06,
+      "loss": 0.2244,
+      "step": 1797
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 10.05634352685722,
+      "learning_rate": 3.1222603889525093e-06,
+      "loss": 0.188,
+      "step": 1798
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 14.762758681652592,
+      "learning_rate": 3.1169677291711774e-06,
+      "loss": 0.1985,
+      "step": 1799
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 10.569747412993385,
+      "learning_rate": 3.11167752688162e-06,
+      "loss": 0.1719,
+      "step": 1800
+    },
+    {
+      "epoch": 2.54,
+      "eval_avg_AUC": 0.764240118126001,
+      "eval_avg_Accuracy": 0.6693468169761273,
+      "eval_avg_Accuracy-right": 0.9362853788965697,
+      "eval_avg_Accuracy-wrong": 0.20388901523766204,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6632694501039103,
+      "eval_last_AUC": 0.7902255750279006,
+      "eval_last_Accuracy": 0.7307692307692307,
+      "eval_last_Accuracy-right": 0.7795095865397157,
+      "eval_last_Accuracy-wrong": 0.6457812144644075,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.683047259724481,
+      "eval_max_AUC": 0.5733454479839537,
+      "eval_max_Accuracy": 0.6339522546419099,
+      "eval_max_Accuracy-right": 0.9853919394808921,
+      "eval_max_Accuracy-wrong": 0.021150784625881283,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5402652358571859,
+      "eval_min_AUC": 0.7811280962786947,
+      "eval_min_Accuracy": 0.728448275862069,
+      "eval_min_Accuracy-right": 0.7629450893439416,
+      "eval_min_Accuracy-wrong": 0.6682965658403457,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6795746127617386,
+      "eval_prod_AUC": 0.7785793632668492,
+      "eval_prod_Accuracy": 0.7031664456233422,
+      "eval_prod_Accuracy-right": 0.6622538150515195,
+      "eval_prod_Accuracy-wrong": 0.7745053445531044,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6750946615501405,
+      "eval_runtime": 668.1843,
+      "eval_samples_per_second": 36.11,
+      "eval_steps_per_second": 1.128,
+      "eval_sum_AUC": 0.5972603983548905,
+      "eval_sum_Accuracy": 0.6349055039787799,
+      "eval_sum_Accuracy-right": 0.9916525368462241,
+      "eval_sum_Accuracy-wrong": 0.01284967022970207,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6350932224474142,
+      "step": 1800
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 13.034933998196697,
+      "learning_rate": 3.1063897889879346e-06,
+      "loss": 0.1998,
+      "step": 1801
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 9.096801649036465,
+      "learning_rate": 3.1011045223909954e-06,
+      "loss": 0.167,
+      "step": 1802
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 12.414478218771391,
+      "learning_rate": 3.0958217339884523e-06,
+      "loss": 0.2017,
+      "step": 1803
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.369661483904768,
+      "learning_rate": 3.0905414306747228e-06,
+      "loss": 0.1769,
+      "step": 1804
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 9.74172916401176,
+      "learning_rate": 3.085263619340978e-06,
+      "loss": 0.1211,
+      "step": 1805
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 11.70299304354774,
+      "learning_rate": 3.0799883068751433e-06,
+      "loss": 0.1528,
+      "step": 1806
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 7.262344875611629,
+      "learning_rate": 3.0747155001618768e-06,
+      "loss": 0.1283,
+      "step": 1807
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 9.495155152802893,
+      "learning_rate": 3.0694452060825676e-06,
+      "loss": 0.167,
+      "step": 1808
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 9.190232862284889,
+      "learning_rate": 3.064177431515326e-06,
+      "loss": 0.1615,
+      "step": 1809
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 8.707222312180141,
+      "learning_rate": 3.0589121833349764e-06,
+      "loss": 0.1421,
+      "step": 1810
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 8.452310216829048,
+      "learning_rate": 3.053649468413043e-06,
+      "loss": 0.1184,
+      "step": 1811
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 11.576695483680806,
+      "learning_rate": 3.048389293617745e-06,
+      "loss": 0.1576,
+      "step": 1812
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 6.989755033025212,
+      "learning_rate": 3.043131665813988e-06,
+      "loss": 0.1046,
+      "step": 1813
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 6.696553578183441,
+      "learning_rate": 3.037876591863352e-06,
+      "loss": 0.1276,
+      "step": 1814
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.546884244619374,
+      "learning_rate": 3.0326240786240827e-06,
+      "loss": 0.1581,
+      "step": 1815
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 11.451965263379922,
+      "learning_rate": 3.0273741329510852e-06,
+      "loss": 0.1719,
+      "step": 1816
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.292177418431594,
+      "learning_rate": 3.022126761695917e-06,
+      "loss": 0.1748,
+      "step": 1817
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 8.211863315505317,
+      "learning_rate": 3.016881971706772e-06,
+      "loss": 0.1382,
+      "step": 1818
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.855424307452257,
+      "learning_rate": 3.011639769828475e-06,
+      "loss": 0.1539,
+      "step": 1819
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.111315961063828,
+      "learning_rate": 3.0064001629024743e-06,
+      "loss": 0.1108,
+      "step": 1820
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.234334054550184,
+      "learning_rate": 3.0011631577668327e-06,
+      "loss": 0.1686,
+      "step": 1821
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 11.825887201367166,
+      "learning_rate": 2.995928761256214e-06,
+      "loss": 0.1715,
+      "step": 1822
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.106163859695766,
+      "learning_rate": 2.990696980201883e-06,
+      "loss": 0.1527,
+      "step": 1823
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 8.79462015474247,
+      "learning_rate": 2.9854678214316875e-06,
+      "loss": 0.1719,
+      "step": 1824
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 10.236440875249786,
+      "learning_rate": 2.980241291770053e-06,
+      "loss": 0.1863,
+      "step": 1825
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 17.63140303004778,
+      "learning_rate": 2.975017398037974e-06,
+      "loss": 0.2203,
+      "step": 1826
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 12.648314209794771,
+      "learning_rate": 2.969796147053006e-06,
+      "loss": 0.1912,
+      "step": 1827
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.610877314718666,
+      "learning_rate": 2.9645775456292546e-06,
+      "loss": 0.184,
+      "step": 1828
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.953679210876116,
+      "learning_rate": 2.9593616005773695e-06,
+      "loss": 0.1423,
+      "step": 1829
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 8.816075738258004,
+      "learning_rate": 2.9541483187045293e-06,
+      "loss": 0.1919,
+      "step": 1830
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.450265693293657,
+      "learning_rate": 2.948937706814442e-06,
+      "loss": 0.1608,
+      "step": 1831
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 6.166075552568411,
+      "learning_rate": 2.9437297717073275e-06,
+      "loss": 0.0922,
+      "step": 1832
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 12.728799555229571,
+      "learning_rate": 2.9385245201799127e-06,
+      "loss": 0.1594,
+      "step": 1833
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 9.465742985640803,
+      "learning_rate": 2.933321959025426e-06,
+      "loss": 0.1813,
+      "step": 1834
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 11.97334962187795,
+      "learning_rate": 2.92812209503358e-06,
+      "loss": 0.1571,
+      "step": 1835
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 8.862884687939331,
+      "learning_rate": 2.9229249349905686e-06,
+      "loss": 0.1295,
+      "step": 1836
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 8.349812453093548,
+      "learning_rate": 2.917730485679058e-06,
+      "loss": 0.1382,
+      "step": 1837
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 6.749903358878047,
+      "learning_rate": 2.9125387538781753e-06,
+      "loss": 0.0951,
+      "step": 1838
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 9.811026729575378,
+      "learning_rate": 2.907349746363502e-06,
+      "loss": 0.1425,
+      "step": 1839
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.861197107923443,
+      "learning_rate": 2.9021634699070646e-06,
+      "loss": 0.1554,
+      "step": 1840
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.304762823029824,
+      "learning_rate": 2.8969799312773263e-06,
+      "loss": 0.1853,
+      "step": 1841
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 6.383597055923346,
+      "learning_rate": 2.8917991372391718e-06,
+      "loss": 0.1171,
+      "step": 1842
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 11.039414401723983,
+      "learning_rate": 2.886621094553912e-06,
+      "loss": 0.1667,
+      "step": 1843
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 6.94621497114141,
+      "learning_rate": 2.8814458099792574e-06,
+      "loss": 0.1184,
+      "step": 1844
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 13.551395458496904,
+      "learning_rate": 2.8762732902693313e-06,
+      "loss": 0.2067,
+      "step": 1845
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.559760738537713,
+      "learning_rate": 2.871103542174637e-06,
+      "loss": 0.2007,
+      "step": 1846
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.042853893833076,
+      "learning_rate": 2.8659365724420687e-06,
+      "loss": 0.176,
+      "step": 1847
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.1724792680764535,
+      "learning_rate": 2.860772387814887e-06,
+      "loss": 0.0905,
+      "step": 1848
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 8.641574637913708,
+      "learning_rate": 2.8556109950327272e-06,
+      "loss": 0.1454,
+      "step": 1849
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 8.725055445873307,
+      "learning_rate": 2.8504524008315697e-06,
+      "loss": 0.1528,
+      "step": 1850
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 7.473503495446864,
+      "learning_rate": 2.8452966119437563e-06,
+      "loss": 0.0866,
+      "step": 1851
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 12.325158834624533,
+      "learning_rate": 2.8401436350979557e-06,
+      "loss": 0.1934,
+      "step": 1852
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.376485393306188,
+      "learning_rate": 2.8349934770191733e-06,
+      "loss": 0.1558,
+      "step": 1853
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 9.567370797614968,
+      "learning_rate": 2.8298461444287305e-06,
+      "loss": 0.1295,
+      "step": 1854
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 7.260894928394038,
+      "learning_rate": 2.8247016440442675e-06,
+      "loss": 0.119,
+      "step": 1855
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 9.778653522794217,
+      "learning_rate": 2.8195599825797233e-06,
+      "loss": 0.1311,
+      "step": 1856
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 9.139564965269917,
+      "learning_rate": 2.814421166745337e-06,
+      "loss": 0.1572,
+      "step": 1857
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 9.471607315047013,
+      "learning_rate": 2.809285203247626e-06,
+      "loss": 0.1363,
+      "step": 1858
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 8.316679930948995,
+      "learning_rate": 2.8041520987893935e-06,
+      "loss": 0.1476,
+      "step": 1859
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.821947042095962,
+      "learning_rate": 2.7990218600697073e-06,
+      "loss": 0.1324,
+      "step": 1860
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 9.449657394885453,
+      "learning_rate": 2.7938944937838924e-06,
+      "loss": 0.1237,
+      "step": 1861
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.46015541713917,
+      "learning_rate": 2.788770006623536e-06,
+      "loss": 0.1272,
+      "step": 1862
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 8.866813047252363,
+      "learning_rate": 2.783648405276452e-06,
+      "loss": 0.1289,
+      "step": 1863
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 8.128306555443599,
+      "learning_rate": 2.7785296964267037e-06,
+      "loss": 0.1049,
+      "step": 1864
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 9.86549879393314,
+      "learning_rate": 2.7734138867545655e-06,
+      "loss": 0.1375,
+      "step": 1865
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.951796968224898,
+      "learning_rate": 2.7683009829365417e-06,
+      "loss": 0.0781,
+      "step": 1866
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 14.53116664645129,
+      "learning_rate": 2.7631909916453283e-06,
+      "loss": 0.2043,
+      "step": 1867
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 11.363614334963833,
+      "learning_rate": 2.7580839195498397e-06,
+      "loss": 0.1487,
+      "step": 1868
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 7.296685911525293,
+      "learning_rate": 2.752979773315163e-06,
+      "loss": 0.1152,
+      "step": 1869
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.425341740647087,
+      "learning_rate": 2.747878559602578e-06,
+      "loss": 0.0948,
+      "step": 1870
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.579999168657471,
+      "learning_rate": 2.7427802850695306e-06,
+      "loss": 0.0741,
+      "step": 1871
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 11.596935501112405,
+      "learning_rate": 2.737684956369635e-06,
+      "loss": 0.1586,
+      "step": 1872
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 8.61857347586662,
+      "learning_rate": 2.732592580152659e-06,
+      "loss": 0.1273,
+      "step": 1873
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 10.680072950551741,
+      "learning_rate": 2.727503163064521e-06,
+      "loss": 0.1802,
+      "step": 1874
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 14.014310467509068,
+      "learning_rate": 2.7224167117472687e-06,
+      "loss": 0.1251,
+      "step": 1875
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.047100742118298,
+      "learning_rate": 2.717333232839088e-06,
+      "loss": 0.1783,
+      "step": 1876
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.358222748920179,
+      "learning_rate": 2.712252732974283e-06,
+      "loss": 0.1161,
+      "step": 1877
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 10.787924061022343,
+      "learning_rate": 2.7071752187832677e-06,
+      "loss": 0.1355,
+      "step": 1878
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 11.001041149987307,
+      "learning_rate": 2.7021006968925613e-06,
+      "loss": 0.2063,
+      "step": 1879
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.696954136219144,
+      "learning_rate": 2.697029173924778e-06,
+      "loss": 0.1748,
+      "step": 1880
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 10.597298813306656,
+      "learning_rate": 2.691960656498621e-06,
+      "loss": 0.192,
+      "step": 1881
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.103880828314004,
+      "learning_rate": 2.686895151228863e-06,
+      "loss": 0.1903,
+      "step": 1882
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 12.900392830303481,
+      "learning_rate": 2.6818326647263547e-06,
+      "loss": 0.2102,
+      "step": 1883
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 14.519263748864603,
+      "learning_rate": 2.6767732035979975e-06,
+      "loss": 0.2588,
+      "step": 1884
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 8.090374473836409,
+      "learning_rate": 2.671716774446758e-06,
+      "loss": 0.1248,
+      "step": 1885
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 8.554240721897587,
+      "learning_rate": 2.6666633838716317e-06,
+      "loss": 0.1184,
+      "step": 1886
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 10.213693972826272,
+      "learning_rate": 2.6616130384676585e-06,
+      "loss": 0.1432,
+      "step": 1887
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 9.757527912245372,
+      "learning_rate": 2.6565657448258964e-06,
+      "loss": 0.1434,
+      "step": 1888
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 9.905308192063403,
+      "learning_rate": 2.651521509533427e-06,
+      "loss": 0.1743,
+      "step": 1889
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 7.547720102252694,
+      "learning_rate": 2.646480339173337e-06,
+      "loss": 0.1095,
+      "step": 1890
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 7.170165517512836,
+      "learning_rate": 2.6414422403247174e-06,
+      "loss": 0.1403,
+      "step": 1891
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 8.337238855549225,
+      "learning_rate": 2.636407219562643e-06,
+      "loss": 0.0961,
+      "step": 1892
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 15.336414631030847,
+      "learning_rate": 2.631375283458178e-06,
+      "loss": 0.2666,
+      "step": 1893
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 7.108946414690598,
+      "learning_rate": 2.626346438578361e-06,
+      "loss": 0.1155,
+      "step": 1894
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 11.779409904177664,
+      "learning_rate": 2.62132069148619e-06,
+      "loss": 0.1592,
+      "step": 1895
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 7.859198510738226,
+      "learning_rate": 2.616298048740626e-06,
+      "loss": 0.1505,
+      "step": 1896
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 11.68336707461125,
+      "learning_rate": 2.6112785168965783e-06,
+      "loss": 0.2283,
+      "step": 1897
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 9.603676724897015,
+      "learning_rate": 2.606262102504895e-06,
+      "loss": 0.1147,
+      "step": 1898
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 9.596570303958858,
+      "learning_rate": 2.6012488121123536e-06,
+      "loss": 0.1501,
+      "step": 1899
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 11.828465280444831,
+      "learning_rate": 2.5962386522616568e-06,
+      "loss": 0.1678,
+      "step": 1900
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 11.358678227503368,
+      "learning_rate": 2.5912316294914232e-06,
+      "loss": 0.142,
+      "step": 1901
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 10.577446995848412,
+      "learning_rate": 2.5862277503361768e-06,
+      "loss": 0.1565,
+      "step": 1902
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 7.950931981171405,
+      "learning_rate": 2.5812270213263326e-06,
+      "loss": 0.1539,
+      "step": 1903
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 7.884663267996451,
+      "learning_rate": 2.5762294489882055e-06,
+      "loss": 0.1444,
+      "step": 1904
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.924639525977193,
+      "learning_rate": 2.57123503984398e-06,
+      "loss": 0.1119,
+      "step": 1905
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 8.5276536593199,
+      "learning_rate": 2.566243800411719e-06,
+      "loss": 0.1565,
+      "step": 1906
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 8.72863320921565,
+      "learning_rate": 2.561255737205347e-06,
+      "loss": 0.1261,
+      "step": 1907
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 11.145168018456879,
+      "learning_rate": 2.556270856734646e-06,
+      "loss": 0.1315,
+      "step": 1908
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 12.48561897448598,
+      "learning_rate": 2.551289165505236e-06,
+      "loss": 0.1957,
+      "step": 1909
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 20.808942542709588,
+      "learning_rate": 2.5463106700185856e-06,
+      "loss": 0.2649,
+      "step": 1910
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 7.458535735883536,
+      "learning_rate": 2.5413353767719805e-06,
+      "loss": 0.112,
+      "step": 1911
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 9.51756112182077,
+      "learning_rate": 2.536363292258543e-06,
+      "loss": 0.1521,
+      "step": 1912
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 8.709498593163685,
+      "learning_rate": 2.5313944229671925e-06,
+      "loss": 0.1202,
+      "step": 1913
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 11.506920993331828,
+      "learning_rate": 2.526428775382661e-06,
+      "loss": 0.1383,
+      "step": 1914
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 6.836743081233571,
+      "learning_rate": 2.5214663559854758e-06,
+      "loss": 0.1145,
+      "step": 1915
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 7.146180629791272,
+      "learning_rate": 2.5165071712519447e-06,
+      "loss": 0.1219,
+      "step": 1916
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 6.642978542424022,
+      "learning_rate": 2.51155122765416e-06,
+      "loss": 0.1184,
+      "step": 1917
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.10548076867044,
+      "learning_rate": 2.506598531659984e-06,
+      "loss": 0.1565,
+      "step": 1918
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.573153980063179,
+      "learning_rate": 2.5016490897330405e-06,
+      "loss": 0.1628,
+      "step": 1919
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 7.412117536111318,
+      "learning_rate": 2.4967029083327003e-06,
+      "loss": 0.1361,
+      "step": 1920
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.101718795614163,
+      "learning_rate": 2.4917599939140884e-06,
+      "loss": 0.1335,
+      "step": 1921
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 8.449627396553131,
+      "learning_rate": 2.486820352928058e-06,
+      "loss": 0.1555,
+      "step": 1922
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 9.639422622055571,
+      "learning_rate": 2.4818839918211963e-06,
+      "loss": 0.1705,
+      "step": 1923
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 36.873576284129875,
+      "learning_rate": 2.4769509170358065e-06,
+      "loss": 0.3184,
+      "step": 1924
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 8.644262645862426,
+      "learning_rate": 2.4720211350099055e-06,
+      "loss": 0.1178,
+      "step": 1925
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 8.36653310411647,
+      "learning_rate": 2.467094652177209e-06,
+      "loss": 0.1567,
+      "step": 1926
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 9.518406580381606,
+      "learning_rate": 2.462171474967133e-06,
+      "loss": 0.116,
+      "step": 1927
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 9.780444856627813,
+      "learning_rate": 2.4572516098047684e-06,
+      "loss": 0.1324,
+      "step": 1928
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 14.180205493929815,
+      "learning_rate": 2.4523350631109004e-06,
+      "loss": 0.2045,
+      "step": 1929
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 10.026309449176285,
+      "learning_rate": 2.4474218413019685e-06,
+      "loss": 0.1201,
+      "step": 1930
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 15.065172930244865,
+      "learning_rate": 2.4425119507900813e-06,
+      "loss": 0.1332,
+      "step": 1931
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 6.197518159942636,
+      "learning_rate": 2.4376053979829938e-06,
+      "loss": 0.0999,
+      "step": 1932
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 8.670582821772705,
+      "learning_rate": 2.4327021892841095e-06,
+      "loss": 0.129,
+      "step": 1933
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 6.891485996833124,
+      "learning_rate": 2.4278023310924676e-06,
+      "loss": 0.1317,
+      "step": 1934
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 7.97301468090986,
+      "learning_rate": 2.4229058298027323e-06,
+      "loss": 0.0712,
+      "step": 1935
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.257165856955314,
+      "learning_rate": 2.418012691805191e-06,
+      "loss": 0.1449,
+      "step": 1936
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 8.358246852023568,
+      "learning_rate": 2.413122923485734e-06,
+      "loss": 0.1337,
+      "step": 1937
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 15.458148790531787,
+      "learning_rate": 2.4082365312258628e-06,
+      "loss": 0.2167,
+      "step": 1938
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 14.434756359865831,
+      "learning_rate": 2.4033535214026625e-06,
+      "loss": 0.2147,
+      "step": 1939
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 9.618075870551374,
+      "learning_rate": 2.3984739003888192e-06,
+      "loss": 0.1465,
+      "step": 1940
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 8.600582581104133,
+      "learning_rate": 2.39359767455258e-06,
+      "loss": 0.1271,
+      "step": 1941
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.970171027055446,
+      "learning_rate": 2.3887248502577727e-06,
+      "loss": 0.1313,
+      "step": 1942
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 11.690717315175087,
+      "learning_rate": 2.3838554338637777e-06,
+      "loss": 0.1852,
+      "step": 1943
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 11.23675443269888,
+      "learning_rate": 2.3789894317255353e-06,
+      "loss": 0.2119,
+      "step": 1944
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.674864185070168,
+      "learning_rate": 2.3741268501935212e-06,
+      "loss": 0.1779,
+      "step": 1945
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 7.149078957075336,
+      "learning_rate": 2.3692676956137585e-06,
+      "loss": 0.1279,
+      "step": 1946
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 9.862752444468342,
+      "learning_rate": 2.364411974327787e-06,
+      "loss": 0.1761,
+      "step": 1947
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 9.538549840490077,
+      "learning_rate": 2.3595596926726735e-06,
+      "loss": 0.1454,
+      "step": 1948
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 10.770965896096097,
+      "learning_rate": 2.354710856980988e-06,
+      "loss": 0.1965,
+      "step": 1949
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 10.999821972084437,
+      "learning_rate": 2.3498654735808107e-06,
+      "loss": 0.1442,
+      "step": 1950
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 9.421264513028135,
+      "learning_rate": 2.3450235487957136e-06,
+      "loss": 0.1986,
+      "step": 1951
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.130629305678671,
+      "learning_rate": 2.3401850889447536e-06,
+      "loss": 0.1096,
+      "step": 1952
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 9.31557017583868,
+      "learning_rate": 2.33535010034247e-06,
+      "loss": 0.176,
+      "step": 1953
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.540503599505865,
+      "learning_rate": 2.3305185892988645e-06,
+      "loss": 0.1519,
+      "step": 1954
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 7.580444314371188,
+      "learning_rate": 2.3256905621194082e-06,
+      "loss": 0.1345,
+      "step": 1955
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 8.211772605220997,
+      "learning_rate": 2.320866025105016e-06,
+      "loss": 0.1412,
+      "step": 1956
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 7.171027964753846,
+      "learning_rate": 2.3160449845520614e-06,
+      "loss": 0.1151,
+      "step": 1957
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.864012619318704,
+      "learning_rate": 2.311227446752342e-06,
+      "loss": 0.0957,
+      "step": 1958
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.294384732813135,
+      "learning_rate": 2.306413417993092e-06,
+      "loss": 0.162,
+      "step": 1959
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.542420819352373,
+      "learning_rate": 2.3016029045569607e-06,
+      "loss": 0.1617,
+      "step": 1960
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 6.389783592731984,
+      "learning_rate": 2.296795912722014e-06,
+      "loss": 0.1432,
+      "step": 1961
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 7.442051708219327,
+      "learning_rate": 2.2919924487617164e-06,
+      "loss": 0.137,
+      "step": 1962
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 7.300951053616588,
+      "learning_rate": 2.2871925189449377e-06,
+      "loss": 0.1323,
+      "step": 1963
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 8.07323617382073,
+      "learning_rate": 2.2823961295359247e-06,
+      "loss": 0.165,
+      "step": 1964
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 7.9225155373766185,
+      "learning_rate": 2.277603286794313e-06,
+      "loss": 0.1458,
+      "step": 1965
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 7.1437752215485215,
+      "learning_rate": 2.2728139969751005e-06,
+      "loss": 0.1199,
+      "step": 1966
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 10.14441691188025,
+      "learning_rate": 2.268028266328655e-06,
+      "loss": 0.142,
+      "step": 1967
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 9.86742282575186,
+      "learning_rate": 2.2632461011006974e-06,
+      "loss": 0.1655,
+      "step": 1968
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 8.228798601901707,
+      "learning_rate": 2.258467507532297e-06,
+      "loss": 0.103,
+      "step": 1969
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 6.690249496747353,
+      "learning_rate": 2.2536924918598563e-06,
+      "loss": 0.1276,
+      "step": 1970
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 7.555764827004685,
+      "learning_rate": 2.2489210603151144e-06,
+      "loss": 0.1547,
+      "step": 1971
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 8.181707492716276,
+      "learning_rate": 2.2441532191251325e-06,
+      "loss": 0.1293,
+      "step": 1972
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 8.384067483981713,
+      "learning_rate": 2.239388974512279e-06,
+      "loss": 0.1388,
+      "step": 1973
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 13.65006367044336,
+      "learning_rate": 2.23462833269424e-06,
+      "loss": 0.1978,
+      "step": 1974
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 11.821231252748818,
+      "learning_rate": 2.22987129988399e-06,
+      "loss": 0.186,
+      "step": 1975
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 7.212791446550921,
+      "learning_rate": 2.225117882289799e-06,
+      "loss": 0.1219,
+      "step": 1976
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 7.8896545620242025,
+      "learning_rate": 2.220368086115214e-06,
+      "loss": 0.1185,
+      "step": 1977
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 6.313148707741496,
+      "learning_rate": 2.2156219175590623e-06,
+      "loss": 0.1268,
+      "step": 1978
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 8.637406927423593,
+      "learning_rate": 2.2108793828154266e-06,
+      "loss": 0.1315,
+      "step": 1979
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 9.721079691858396,
+      "learning_rate": 2.2061404880736638e-06,
+      "loss": 0.1688,
+      "step": 1980
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 12.495339431159342,
+      "learning_rate": 2.201405239518363e-06,
+      "loss": 0.1948,
+      "step": 1981
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 11.801355256971291,
+      "learning_rate": 2.1966736433293657e-06,
+      "loss": 0.1646,
+      "step": 1982
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 16.49928534020883,
+      "learning_rate": 2.1919457056817406e-06,
+      "loss": 0.1831,
+      "step": 1983
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 10.353121505612982,
+      "learning_rate": 2.1872214327457856e-06,
+      "loss": 0.1299,
+      "step": 1984
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 7.179144940448414,
+      "learning_rate": 2.182500830687016e-06,
+      "loss": 0.1012,
+      "step": 1985
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 7.74542754298966,
+      "learning_rate": 2.1777839056661555e-06,
+      "loss": 0.1206,
+      "step": 1986
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 11.963890332481387,
+      "learning_rate": 2.173070663839125e-06,
+      "loss": 0.1613,
+      "step": 1987
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 11.573390310201928,
+      "learning_rate": 2.1683611113570446e-06,
+      "loss": 0.1107,
+      "step": 1988
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 8.153529587971578,
+      "learning_rate": 2.1636552543662187e-06,
+      "loss": 0.1462,
+      "step": 1989
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 15.240802101937277,
+      "learning_rate": 2.1589530990081237e-06,
+      "loss": 0.178,
+      "step": 1990
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.427408431263819,
+      "learning_rate": 2.1542546514194103e-06,
+      "loss": 0.0984,
+      "step": 1991
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.93273826392774,
+      "learning_rate": 2.149559917731889e-06,
+      "loss": 0.1045,
+      "step": 1992
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 8.966709874278905,
+      "learning_rate": 2.1448689040725237e-06,
+      "loss": 0.1188,
+      "step": 1993
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 8.093498764756616,
+      "learning_rate": 2.1401816165634197e-06,
+      "loss": 0.1172,
+      "step": 1994
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 10.013500748394748,
+      "learning_rate": 2.1354980613218244e-06,
+      "loss": 0.1464,
+      "step": 1995
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 6.763711927216121,
+      "learning_rate": 2.1308182444601126e-06,
+      "loss": 0.1066,
+      "step": 1996
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 8.339160281624952,
+      "learning_rate": 2.126142172085781e-06,
+      "loss": 0.1322,
+      "step": 1997
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 10.115549164536688,
+      "learning_rate": 2.1214698503014358e-06,
+      "loss": 0.1633,
+      "step": 1998
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 7.348778318559917,
+      "learning_rate": 2.1168012852047936e-06,
+      "loss": 0.1119,
+      "step": 1999
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 16.944068120646698,
+      "learning_rate": 2.112136482888663e-06,
+      "loss": 0.1694,
+      "step": 2000
+    },
+    {
+      "epoch": 2.82,
+      "eval_avg_AUC": 0.7624768000366875,
+      "eval_avg_Accuracy": 0.6572861405835544,
+      "eval_avg_Accuracy-right": 0.9455458458327899,
+      "eval_avg_Accuracy-wrong": 0.15465089833977713,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.656654174340923,
+      "eval_last_AUC": 0.7891685044752432,
+      "eval_last_Accuracy": 0.7292771883289124,
+      "eval_last_Accuracy-right": 0.8100952132515977,
+      "eval_last_Accuracy-wrong": 0.5883556970661815,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6677202444929105,
+      "eval_max_AUC": 0.5551535828746488,
+      "eval_max_Accuracy": 0.633040450928382,
+      "eval_max_Accuracy-right": 0.9905438893961133,
+      "eval_max_Accuracy-wrong": 0.00966568114623607,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5314323592750985,
+      "eval_min_AUC": 0.7789547919112831,
+      "eval_min_Accuracy": 0.7274121352785146,
+      "eval_min_Accuracy-right": 0.7962697274031564,
+      "eval_min_Accuracy-wrong": 0.6073459176711394,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6653990971421216,
+      "eval_prod_AUC": 0.7770646228342784,
+      "eval_prod_Accuracy": 0.7132377320954907,
+      "eval_prod_Accuracy-right": 0.7168383983305073,
+      "eval_prod_Accuracy-wrong": 0.7069592904252899,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.665843784045975,
+      "eval_runtime": 667.9195,
+      "eval_samples_per_second": 36.124,
+      "eval_steps_per_second": 1.129,
+      "eval_sum_AUC": 0.5784173406171531,
+      "eval_sum_Accuracy": 0.6348640583554377,
+      "eval_sum_Accuracy-right": 0.994848050084779,
+      "eval_sum_Accuracy-wrong": 0.007163975437798499,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6186408123966357,
+      "step": 2000
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 7.454543380585548,
+      "learning_rate": 2.1074754494409462e-06,
+      "loss": 0.1026,
+      "step": 2001
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 7.730697117214318,
+      "learning_rate": 2.1028181909446244e-06,
+      "loss": 0.1251,
+      "step": 2002
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.233758509007847,
+      "learning_rate": 2.098164713477756e-06,
+      "loss": 0.2363,
+      "step": 2003
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 7.98469203947412,
+      "learning_rate": 2.093515023113458e-06,
+      "loss": 0.1201,
+      "step": 2004
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 7.611591199914047,
+      "learning_rate": 2.0888691259199097e-06,
+      "loss": 0.1315,
+      "step": 2005
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 7.718503372420222,
+      "learning_rate": 2.0842270279603403e-06,
+      "loss": 0.1086,
+      "step": 2006
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 9.648102024730967,
+      "learning_rate": 2.079588735293021e-06,
+      "loss": 0.0841,
+      "step": 2007
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 10.644000934386682,
+      "learning_rate": 2.074954253971251e-06,
+      "loss": 0.1322,
+      "step": 2008
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 8.785112237942164,
+      "learning_rate": 2.070323590043364e-06,
+      "loss": 0.1373,
+      "step": 2009
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.181162076311972,
+      "learning_rate": 2.065696749552707e-06,
+      "loss": 0.1653,
+      "step": 2010
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 12.475116431368722,
+      "learning_rate": 2.061073738537635e-06,
+      "loss": 0.1711,
+      "step": 2011
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 7.3404002224513,
+      "learning_rate": 2.0564545630315098e-06,
+      "loss": 0.13,
+      "step": 2012
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 7.300389348268909,
+      "learning_rate": 2.051839229062686e-06,
+      "loss": 0.1318,
+      "step": 2013
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 11.431763627240922,
+      "learning_rate": 2.0472277426545054e-06,
+      "loss": 0.2042,
+      "step": 2014
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 9.193744179066567,
+      "learning_rate": 2.042620109825285e-06,
+      "loss": 0.1094,
+      "step": 2015
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 10.283189702702902,
+      "learning_rate": 2.0380163365883188e-06,
+      "loss": 0.1241,
+      "step": 2016
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 9.456186442940572,
+      "learning_rate": 2.0334164289518556e-06,
+      "loss": 0.1346,
+      "step": 2017
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 8.177975080301406,
+      "learning_rate": 2.028820392919106e-06,
+      "loss": 0.1195,
+      "step": 2018
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 10.6867938177069,
+      "learning_rate": 2.0242282344882257e-06,
+      "loss": 0.1779,
+      "step": 2019
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 8.390545898883135,
+      "learning_rate": 2.019639959652312e-06,
+      "loss": 0.1139,
+      "step": 2020
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 6.575222668599122,
+      "learning_rate": 2.015055574399388e-06,
+      "loss": 0.0825,
+      "step": 2021
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 12.512797359297302,
+      "learning_rate": 2.0104750847124075e-06,
+      "loss": 0.1554,
+      "step": 2022
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 7.074193997682048,
+      "learning_rate": 2.0058984965692317e-06,
+      "loss": 0.1201,
+      "step": 2023
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 10.782656354327477,
+      "learning_rate": 2.0013258159426434e-06,
+      "loss": 0.1562,
+      "step": 2024
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 10.541718035138985,
+      "learning_rate": 1.996757048800312e-06,
+      "loss": 0.1438,
+      "step": 2025
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 7.736481263572269,
+      "learning_rate": 1.9921922011048065e-06,
+      "loss": 0.098,
+      "step": 2026
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 7.84795693765476,
+      "learning_rate": 1.987631278813582e-06,
+      "loss": 0.1058,
+      "step": 2027
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 13.493646022723384,
+      "learning_rate": 1.983074287878965e-06,
+      "loss": 0.1924,
+      "step": 2028
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 8.01412528645441,
+      "learning_rate": 1.9785212342481556e-06,
+      "loss": 0.1285,
+      "step": 2029
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 10.617289242119252,
+      "learning_rate": 1.9739721238632154e-06,
+      "loss": 0.1813,
+      "step": 2030
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 7.016454148888036,
+      "learning_rate": 1.969426962661059e-06,
+      "loss": 0.1387,
+      "step": 2031
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 10.010842355560039,
+      "learning_rate": 1.9648857565734447e-06,
+      "loss": 0.1169,
+      "step": 2032
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 7.3458576402821825,
+      "learning_rate": 1.9603485115269743e-06,
+      "loss": 0.1152,
+      "step": 2033
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 9.520592893854312,
+      "learning_rate": 1.955815233443072e-06,
+      "loss": 0.1826,
+      "step": 2034
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 10.370168724309517,
+      "learning_rate": 1.951285928237996e-06,
+      "loss": 0.1581,
+      "step": 2035
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 6.320658304328653,
+      "learning_rate": 1.946760601822809e-06,
+      "loss": 0.0931,
+      "step": 2036
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 8.000819758479938,
+      "learning_rate": 1.9422392601033896e-06,
+      "loss": 0.1328,
+      "step": 2037
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 10.077353379182258,
+      "learning_rate": 1.937721908980407e-06,
+      "loss": 0.1508,
+      "step": 2038
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.688883712215641,
+      "learning_rate": 1.9332085543493328e-06,
+      "loss": 0.0864,
+      "step": 2039
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 9.103127138590027,
+      "learning_rate": 1.928699202100411e-06,
+      "loss": 0.1255,
+      "step": 2040
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 9.734554979988777,
+      "learning_rate": 1.9241938581186766e-06,
+      "loss": 0.103,
+      "step": 2041
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 8.892637409982045,
+      "learning_rate": 1.9196925282839195e-06,
+      "loss": 0.1371,
+      "step": 2042
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 8.484482618722769,
+      "learning_rate": 1.9151952184707014e-06,
+      "loss": 0.1416,
+      "step": 2043
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 6.560240132723988,
+      "learning_rate": 1.910701934548329e-06,
+      "loss": 0.08,
+      "step": 2044
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.939529615703206,
+      "learning_rate": 1.906212682380862e-06,
+      "loss": 0.0961,
+      "step": 2045
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 13.035369010352479,
+      "learning_rate": 1.9017274678270948e-06,
+      "loss": 0.1914,
+      "step": 2046
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 9.260216182154364,
+      "learning_rate": 1.8972462967405531e-06,
+      "loss": 0.1322,
+      "step": 2047
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 9.190465837366212,
+      "learning_rate": 1.8927691749694871e-06,
+      "loss": 0.1495,
+      "step": 2048
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 8.006983715392067,
+      "learning_rate": 1.888296108356858e-06,
+      "loss": 0.1075,
+      "step": 2049
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 13.40044020252234,
+      "learning_rate": 1.883827102740341e-06,
+      "loss": 0.1802,
+      "step": 2050
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 11.59554704797835,
+      "learning_rate": 1.879362163952303e-06,
+      "loss": 0.1599,
+      "step": 2051
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 9.631586794711819,
+      "learning_rate": 1.8749012978198144e-06,
+      "loss": 0.1289,
+      "step": 2052
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 13.75896724447601,
+      "learning_rate": 1.8704445101646207e-06,
+      "loss": 0.136,
+      "step": 2053
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 11.37345710654823,
+      "learning_rate": 1.8659918068031506e-06,
+      "loss": 0.1451,
+      "step": 2054
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.85712293068289,
+      "learning_rate": 1.8615431935464984e-06,
+      "loss": 0.1518,
+      "step": 2055
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 9.233878961577336,
+      "learning_rate": 1.8570986762004246e-06,
+      "loss": 0.1318,
+      "step": 2056
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 10.44890566280647,
+      "learning_rate": 1.8526582605653376e-06,
+      "loss": 0.1484,
+      "step": 2057
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.706574461250501,
+      "learning_rate": 1.848221952436305e-06,
+      "loss": 0.0714,
+      "step": 2058
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 7.497331395457189,
+      "learning_rate": 1.8437897576030195e-06,
+      "loss": 0.1191,
+      "step": 2059
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 11.751349914766747,
+      "learning_rate": 1.8393616818498173e-06,
+      "loss": 0.1608,
+      "step": 2060
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 8.853903180300453,
+      "learning_rate": 1.8349377309556487e-06,
+      "loss": 0.1173,
+      "step": 2061
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 12.15218013949701,
+      "learning_rate": 1.8305179106940896e-06,
+      "loss": 0.1401,
+      "step": 2062
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 11.135478780447036,
+      "learning_rate": 1.8261022268333205e-06,
+      "loss": 0.1714,
+      "step": 2063
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 10.361752004649604,
+      "learning_rate": 1.8216906851361265e-06,
+      "loss": 0.1107,
+      "step": 2064
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 8.046858209551303,
+      "learning_rate": 1.8172832913598814e-06,
+      "loss": 0.1451,
+      "step": 2065
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 6.585724538347267,
+      "learning_rate": 1.8128800512565514e-06,
+      "loss": 0.0865,
+      "step": 2066
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 7.827995858943193,
+      "learning_rate": 1.8084809705726814e-06,
+      "loss": 0.1027,
+      "step": 2067
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 9.449447563657634,
+      "learning_rate": 1.8040860550493805e-06,
+      "loss": 0.1273,
+      "step": 2068
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 12.940481367487457,
+      "learning_rate": 1.7996953104223357e-06,
+      "loss": 0.1548,
+      "step": 2069
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 12.11485100567967,
+      "learning_rate": 1.7953087424217764e-06,
+      "loss": 0.176,
+      "step": 2070
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 14.585401672739456,
+      "learning_rate": 1.7909263567724917e-06,
+      "loss": 0.1068,
+      "step": 2071
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 7.970140280477049,
+      "learning_rate": 1.786548159193805e-06,
+      "loss": 0.1077,
+      "step": 2072
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 13.824690786926826,
+      "learning_rate": 1.7821741553995798e-06,
+      "loss": 0.1376,
+      "step": 2073
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 10.589215165751035,
+      "learning_rate": 1.7778043510981997e-06,
+      "loss": 0.1155,
+      "step": 2074
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 7.2906632570859715,
+      "learning_rate": 1.7734387519925783e-06,
+      "loss": 0.1635,
+      "step": 2075
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.8414026817470734,
+      "learning_rate": 1.7690773637801295e-06,
+      "loss": 0.0814,
+      "step": 2076
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 6.8689270681921295,
+      "learning_rate": 1.7647201921527802e-06,
+      "loss": 0.1085,
+      "step": 2077
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 12.455066475547923,
+      "learning_rate": 1.7603672427969471e-06,
+      "loss": 0.1533,
+      "step": 2078
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 11.684280310989184,
+      "learning_rate": 1.7560185213935427e-06,
+      "loss": 0.1368,
+      "step": 2079
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 9.717465289750898,
+      "learning_rate": 1.7516740336179595e-06,
+      "loss": 0.1385,
+      "step": 2080
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 8.206664484021074,
+      "learning_rate": 1.7473337851400663e-06,
+      "loss": 0.1233,
+      "step": 2081
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 9.678368875503557,
+      "learning_rate": 1.7429977816241938e-06,
+      "loss": 0.1138,
+      "step": 2082
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 9.603991042007172,
+      "learning_rate": 1.7386660287291395e-06,
+      "loss": 0.1359,
+      "step": 2083
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 9.350733954926785,
+      "learning_rate": 1.7343385321081514e-06,
+      "loss": 0.1386,
+      "step": 2084
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 9.382654379748935,
+      "learning_rate": 1.7300152974089196e-06,
+      "loss": 0.1425,
+      "step": 2085
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 14.19529669253149,
+      "learning_rate": 1.7256963302735752e-06,
+      "loss": 0.2356,
+      "step": 2086
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 7.514377595626204,
+      "learning_rate": 1.7213816363386798e-06,
+      "loss": 0.1281,
+      "step": 2087
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 11.490369523213733,
+      "learning_rate": 1.7170712212352187e-06,
+      "loss": 0.1578,
+      "step": 2088
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 11.681103303832883,
+      "learning_rate": 1.7127650905885896e-06,
+      "loss": 0.1532,
+      "step": 2089
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 7.6536891299576855,
+      "learning_rate": 1.7084632500186016e-06,
+      "loss": 0.1125,
+      "step": 2090
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 15.422564368634989,
+      "learning_rate": 1.7041657051394645e-06,
+      "loss": 0.1367,
+      "step": 2091
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 7.6883343967679405,
+      "learning_rate": 1.699872461559784e-06,
+      "loss": 0.0991,
+      "step": 2092
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 10.009353535368486,
+      "learning_rate": 1.6955835248825453e-06,
+      "loss": 0.1605,
+      "step": 2093
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 13.575613609364757,
+      "learning_rate": 1.6912989007051224e-06,
+      "loss": 0.1143,
+      "step": 2094
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.48667233237705,
+      "learning_rate": 1.6870185946192518e-06,
+      "loss": 0.0721,
+      "step": 2095
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 8.604139471002835,
+      "learning_rate": 1.6827426122110412e-06,
+      "loss": 0.1099,
+      "step": 2096
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 9.36033585218402,
+      "learning_rate": 1.6784709590609539e-06,
+      "loss": 0.142,
+      "step": 2097
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 12.762691200398761,
+      "learning_rate": 1.6742036407438034e-06,
+      "loss": 0.1498,
+      "step": 2098
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 11.376884058838588,
+      "learning_rate": 1.6699406628287423e-06,
+      "loss": 0.1594,
+      "step": 2099
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.844272733337939,
+      "learning_rate": 1.6656820308792636e-06,
+      "loss": 0.1085,
+      "step": 2100
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 11.462351450826803,
+      "learning_rate": 1.6614277504531867e-06,
+      "loss": 0.1624,
+      "step": 2101
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 10.612206934863352,
+      "learning_rate": 1.657177827102653e-06,
+      "loss": 0.1244,
+      "step": 2102
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 11.384204587925291,
+      "learning_rate": 1.6529322663741127e-06,
+      "loss": 0.1482,
+      "step": 2103
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 13.222294921184915,
+      "learning_rate": 1.648691073808329e-06,
+      "loss": 0.1333,
+      "step": 2104
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 10.130214611732262,
+      "learning_rate": 1.644454254940362e-06,
+      "loss": 0.179,
+      "step": 2105
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 8.12595228277028,
+      "learning_rate": 1.6402218152995609e-06,
+      "loss": 0.1119,
+      "step": 2106
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 7.006519857966507,
+      "learning_rate": 1.6359937604095632e-06,
+      "loss": 0.1255,
+      "step": 2107
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 7.560380512808602,
+      "learning_rate": 1.6317700957882832e-06,
+      "loss": 0.1036,
+      "step": 2108
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 8.02888132060872,
+      "learning_rate": 1.6275508269479068e-06,
+      "loss": 0.1676,
+      "step": 2109
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 11.159513040261919,
+      "learning_rate": 1.6233359593948777e-06,
+      "loss": 0.1445,
+      "step": 2110
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 9.10054089940727,
+      "learning_rate": 1.6191254986299044e-06,
+      "loss": 0.1296,
+      "step": 2111
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 12.67247306583035,
+      "learning_rate": 1.6149194501479343e-06,
+      "loss": 0.1256,
+      "step": 2112
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 7.813620467434575,
+      "learning_rate": 1.610717819438164e-06,
+      "loss": 0.1543,
+      "step": 2113
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 12.537185683320704,
+      "learning_rate": 1.6065206119840216e-06,
+      "loss": 0.1449,
+      "step": 2114
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 7.083088201651096,
+      "learning_rate": 1.6023278332631653e-06,
+      "loss": 0.0948,
+      "step": 2115
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 11.121855143786968,
+      "learning_rate": 1.598139488747467e-06,
+      "loss": 0.1606,
+      "step": 2116
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 8.907986157041563,
+      "learning_rate": 1.5939555839030195e-06,
+      "loss": 0.1553,
+      "step": 2117
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 14.891725406567215,
+      "learning_rate": 1.5897761241901138e-06,
+      "loss": 0.1841,
+      "step": 2118
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 9.348992391680904,
+      "learning_rate": 1.5856011150632488e-06,
+      "loss": 0.1593,
+      "step": 2119
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 6.971297803808772,
+      "learning_rate": 1.5814305619711057e-06,
+      "loss": 0.1057,
+      "step": 2120
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 8.457513336541316,
+      "learning_rate": 1.5772644703565564e-06,
+      "loss": 0.1219,
+      "step": 2121
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 12.521801662264629,
+      "learning_rate": 1.5731028456566499e-06,
+      "loss": 0.1561,
+      "step": 2122
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 8.2666981394845,
+      "learning_rate": 1.5689456933026004e-06,
+      "loss": 0.137,
+      "step": 2123
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 10.746275585395676,
+      "learning_rate": 1.5647930187197914e-06,
+      "loss": 0.1284,
+      "step": 2124
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 9.02418620896233,
+      "learning_rate": 1.5606448273277592e-06,
+      "loss": 0.1437,
+      "step": 2125
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 22.317369810298935,
+      "learning_rate": 1.5565011245401928e-06,
+      "loss": 0.2095,
+      "step": 2126
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 8.009535860418744,
+      "learning_rate": 1.5523619157649167e-06,
+      "loss": 0.1592,
+      "step": 2127
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.678611570807112,
+      "learning_rate": 1.5482272064038984e-06,
+      "loss": 0.0748,
+      "step": 2128
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 6.852201122848066,
+      "learning_rate": 1.5440970018532237e-06,
+      "loss": 0.0956,
+      "step": 2129
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.983902183442478,
+      "learning_rate": 1.539971307503112e-06,
+      "loss": 0.0955,
+      "step": 2130
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 6.112518692994072,
+      "learning_rate": 1.535850128737884e-06,
+      "loss": 0.0729,
+      "step": 2131
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 5.484034364294167,
+      "learning_rate": 1.531733470935976e-06,
+      "loss": 0.082,
+      "step": 2132
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 9.313422545173568,
+      "learning_rate": 1.5276213394699184e-06,
+      "loss": 0.1395,
+      "step": 2133
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 6.466431698354635,
+      "learning_rate": 1.5235137397063394e-06,
+      "loss": 0.0687,
+      "step": 2134
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 3.1139218475971018,
+      "learning_rate": 1.5194106770059452e-06,
+      "loss": 0.0471,
+      "step": 2135
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 7.039624892077299,
+      "learning_rate": 1.5153121567235334e-06,
+      "loss": 0.0887,
+      "step": 2136
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 2.644428780207753,
+      "learning_rate": 1.511218184207961e-06,
+      "loss": 0.041,
+      "step": 2137
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 5.759312312983082,
+      "learning_rate": 1.5071287648021577e-06,
+      "loss": 0.0695,
+      "step": 2138
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 6.076034960766147,
+      "learning_rate": 1.5030439038431054e-06,
+      "loss": 0.1128,
+      "step": 2139
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 2.9608286550108662,
+      "learning_rate": 1.4989636066618413e-06,
+      "loss": 0.0446,
+      "step": 2140
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 4.40436218795221,
+      "learning_rate": 1.4948878785834453e-06,
+      "loss": 0.0651,
+      "step": 2141
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 3.7977755318085404,
+      "learning_rate": 1.4908167249270334e-06,
+      "loss": 0.0566,
+      "step": 2142
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 5.436680502538212,
+      "learning_rate": 1.4867501510057548e-06,
+      "loss": 0.0852,
+      "step": 2143
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 4.032630492559908,
+      "learning_rate": 1.4826881621267753e-06,
+      "loss": 0.064,
+      "step": 2144
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 4.177858314023339,
+      "learning_rate": 1.4786307635912839e-06,
+      "loss": 0.0712,
+      "step": 2145
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 3.5685347307421416,
+      "learning_rate": 1.4745779606944716e-06,
+      "loss": 0.0722,
+      "step": 2146
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 3.483034141481542,
+      "learning_rate": 1.4705297587255425e-06,
+      "loss": 0.0587,
+      "step": 2147
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 3.999470814411376,
+      "learning_rate": 1.4664861629676846e-06,
+      "loss": 0.0709,
+      "step": 2148
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 4.472266279685734,
+      "learning_rate": 1.4624471786980826e-06,
+      "loss": 0.0676,
+      "step": 2149
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 6.1012604362815015,
+      "learning_rate": 1.458412811187897e-06,
+      "loss": 0.0821,
+      "step": 2150
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 5.9551887575945575,
+      "learning_rate": 1.4543830657022683e-06,
+      "loss": 0.0658,
+      "step": 2151
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.3743637477502766,
+      "learning_rate": 1.4503579475002993e-06,
+      "loss": 0.0524,
+      "step": 2152
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 6.1763694200058525,
+      "learning_rate": 1.4463374618350617e-06,
+      "loss": 0.0645,
+      "step": 2153
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 4.201190754997552,
+      "learning_rate": 1.4423216139535735e-06,
+      "loss": 0.0603,
+      "step": 2154
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.262356142099759,
+      "learning_rate": 1.438310409096806e-06,
+      "loss": 0.0518,
+      "step": 2155
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 6.028058220077656,
+      "learning_rate": 1.4343038524996645e-06,
+      "loss": 0.072,
+      "step": 2156
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 4.499898657828777,
+      "learning_rate": 1.4303019493909947e-06,
+      "loss": 0.0623,
+      "step": 2157
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 4.228350000882751,
+      "learning_rate": 1.4263047049935658e-06,
+      "loss": 0.0634,
+      "step": 2158
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 4.123797882874982,
+      "learning_rate": 1.4223121245240674e-06,
+      "loss": 0.0729,
+      "step": 2159
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 6.888639150695895,
+      "learning_rate": 1.4183242131931041e-06,
+      "loss": 0.1086,
+      "step": 2160
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 4.685587910918083,
+      "learning_rate": 1.4143409762051829e-06,
+      "loss": 0.0668,
+      "step": 2161
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 5.391967408684052,
+      "learning_rate": 1.410362418758715e-06,
+      "loss": 0.0939,
+      "step": 2162
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 3.6225799712579914,
+      "learning_rate": 1.4063885460459987e-06,
+      "loss": 0.0542,
+      "step": 2163
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 5.329334687090934,
+      "learning_rate": 1.4024193632532279e-06,
+      "loss": 0.0777,
+      "step": 2164
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 9.225393128089404,
+      "learning_rate": 1.3984548755604655e-06,
+      "loss": 0.0799,
+      "step": 2165
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 4.461759465244987,
+      "learning_rate": 1.3944950881416541e-06,
+      "loss": 0.0642,
+      "step": 2166
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.4394014361986343,
+      "learning_rate": 1.3905400061645968e-06,
+      "loss": 0.0532,
+      "step": 2167
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 4.315808038217923,
+      "learning_rate": 1.3865896347909614e-06,
+      "loss": 0.0672,
+      "step": 2168
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 4.26791022615535,
+      "learning_rate": 1.3826439791762602e-06,
+      "loss": 0.0667,
+      "step": 2169
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 4.349133357965395,
+      "learning_rate": 1.378703044469863e-06,
+      "loss": 0.0622,
+      "step": 2170
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.49651112706006,
+      "learning_rate": 1.3747668358149658e-06,
+      "loss": 0.0474,
+      "step": 2171
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.3764705922255187,
+      "learning_rate": 1.3708353583486055e-06,
+      "loss": 0.0544,
+      "step": 2172
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 6.402368405917794,
+      "learning_rate": 1.3669086172016387e-06,
+      "loss": 0.072,
+      "step": 2173
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 4.075472211653643,
+      "learning_rate": 1.362986617498745e-06,
+      "loss": 0.0579,
+      "step": 2174
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 3.1520672729332744,
+      "learning_rate": 1.359069364358414e-06,
+      "loss": 0.0588,
+      "step": 2175
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 6.061107627345443,
+      "learning_rate": 1.3551568628929434e-06,
+      "loss": 0.0762,
+      "step": 2176
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 3.1127359757661686,
+      "learning_rate": 1.3512491182084236e-06,
+      "loss": 0.0384,
+      "step": 2177
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 12.732811049821692,
+      "learning_rate": 1.3473461354047434e-06,
+      "loss": 0.0755,
+      "step": 2178
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 4.405037286069489,
+      "learning_rate": 1.343447919575575e-06,
+      "loss": 0.0747,
+      "step": 2179
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 4.4912080883231695,
+      "learning_rate": 1.3395544758083646e-06,
+      "loss": 0.0745,
+      "step": 2180
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 3.7166103292445527,
+      "learning_rate": 1.335665809184341e-06,
+      "loss": 0.0524,
+      "step": 2181
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 5.253595985941572,
+      "learning_rate": 1.331781924778488e-06,
+      "loss": 0.052,
+      "step": 2182
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 4.1611260079825225,
+      "learning_rate": 1.3279028276595545e-06,
+      "loss": 0.0421,
+      "step": 2183
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 5.732182676021725,
+      "learning_rate": 1.3240285228900373e-06,
+      "loss": 0.0822,
+      "step": 2184
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 5.398090825256094,
+      "learning_rate": 1.320159015526183e-06,
+      "loss": 0.0705,
+      "step": 2185
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 5.18616918414433,
+      "learning_rate": 1.3162943106179748e-06,
+      "loss": 0.0724,
+      "step": 2186
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 4.166469738095059,
+      "learning_rate": 1.312434413209131e-06,
+      "loss": 0.0568,
+      "step": 2187
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 7.0907160331943215,
+      "learning_rate": 1.3085793283370908e-06,
+      "loss": 0.1589,
+      "step": 2188
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 2.935323888410141,
+      "learning_rate": 1.3047290610330176e-06,
+      "loss": 0.0424,
+      "step": 2189
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 3.8226216036023857,
+      "learning_rate": 1.3008836163217842e-06,
+      "loss": 0.0558,
+      "step": 2190
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 3.717450943128668,
+      "learning_rate": 1.2970429992219714e-06,
+      "loss": 0.0598,
+      "step": 2191
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 8.191840629669882,
+      "learning_rate": 1.2932072147458592e-06,
+      "loss": 0.1044,
+      "step": 2192
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 5.600773469480212,
+      "learning_rate": 1.289376267899422e-06,
+      "loss": 0.0768,
+      "step": 2193
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 5.567022621723898,
+      "learning_rate": 1.285550163682317e-06,
+      "loss": 0.0686,
+      "step": 2194
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 5.264114317205964,
+      "learning_rate": 1.2817289070878853e-06,
+      "loss": 0.0972,
+      "step": 2195
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 8.655026327618923,
+      "learning_rate": 1.2779125031031413e-06,
+      "loss": 0.1011,
+      "step": 2196
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 4.082769054979913,
+      "learning_rate": 1.2741009567087619e-06,
+      "loss": 0.058,
+      "step": 2197
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 3.775704356143737,
+      "learning_rate": 1.2702942728790897e-06,
+      "loss": 0.0535,
+      "step": 2198
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 3.256980107715356,
+      "learning_rate": 1.266492456582119e-06,
+      "loss": 0.0496,
+      "step": 2199
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 4.555534291707936,
+      "learning_rate": 1.262695512779493e-06,
+      "loss": 0.0546,
+      "step": 2200
+    },
+    {
+      "epoch": 3.1,
+      "eval_avg_AUC": 0.7794433856822651,
+      "eval_avg_Accuracy": 0.6601044429708223,
+      "eval_avg_Accuracy-right": 0.9475674970653449,
+      "eval_avg_Accuracy-wrong": 0.15885831248578577,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6624017000468806,
+      "eval_last_AUC": 0.8004563587662588,
+      "eval_last_Accuracy": 0.7387682360742706,
+      "eval_last_Accuracy-right": 0.8127037954871528,
+      "eval_last_Accuracy-wrong": 0.609847623379577,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6778412366233811,
+      "eval_max_AUC": 0.5744885381228098,
+      "eval_max_Accuracy": 0.6350298408488063,
+      "eval_max_Accuracy-right": 0.9931524716316682,
+      "eval_max_Accuracy-wrong": 0.010575392312940642,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5436460591282469,
+      "eval_min_AUC": 0.7938153085511692,
+      "eval_min_Accuracy": 0.7390998010610079,
+      "eval_min_Accuracy-right": 0.8043563323333768,
+      "eval_min_Accuracy-wrong": 0.6253127132135547,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6764654108423312,
+      "eval_prod_AUC": 0.7923313362778415,
+      "eval_prod_Accuracy": 0.7330487400530504,
+      "eval_prod_Accuracy-right": 0.7634015912351637,
+      "eval_prod_Accuracy-wrong": 0.6801228110075052,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6727281670002327,
+      "eval_runtime": 667.1775,
+      "eval_samples_per_second": 36.164,
+      "eval_steps_per_second": 1.13,
+      "eval_sum_AUC": 0.5895669198787048,
+      "eval_sum_Accuracy": 0.6346568302387268,
+      "eval_sum_Accuracy-right": 0.994326333637668,
+      "eval_sum_Accuracy-wrong": 0.007505117125312713,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.635243072060077,
+      "step": 2200
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 4.5420453106181355,
+      "learning_rate": 1.2589034464264933e-06,
+      "loss": 0.0643,
+      "step": 2201
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 8.244039125030884,
+      "learning_rate": 1.255116262472038e-06,
+      "loss": 0.0758,
+      "step": 2202
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 10.81339471506414,
+      "learning_rate": 1.2513339658586731e-06,
+      "loss": 0.0861,
+      "step": 2203
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 4.8408519856239325,
+      "learning_rate": 1.2475565615225683e-06,
+      "loss": 0.0684,
+      "step": 2204
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 4.004095680179641,
+      "learning_rate": 1.2437840543935025e-06,
+      "loss": 0.0511,
+      "step": 2205
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 8.04612287644894,
+      "learning_rate": 1.2400164493948713e-06,
+      "loss": 0.0755,
+      "step": 2206
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 3.8017233954156873,
+      "learning_rate": 1.236253751443665e-06,
+      "loss": 0.0683,
+      "step": 2207
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 7.524513571030522,
+      "learning_rate": 1.2324959654504747e-06,
+      "loss": 0.121,
+      "step": 2208
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.1214649719860468,
+      "learning_rate": 1.2287430963194807e-06,
+      "loss": 0.0424,
+      "step": 2209
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.7336324882955725,
+      "learning_rate": 1.2249951489484468e-06,
+      "loss": 0.0482,
+      "step": 2210
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.389383178418963,
+      "learning_rate": 1.2212521282287093e-06,
+      "loss": 0.0516,
+      "step": 2211
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 5.473451774942144,
+      "learning_rate": 1.217514039045179e-06,
+      "loss": 0.0548,
+      "step": 2212
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 5.5500686037838305,
+      "learning_rate": 1.2137808862763301e-06,
+      "loss": 0.0877,
+      "step": 2213
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 4.830463667481979,
+      "learning_rate": 1.2100526747941954e-06,
+      "loss": 0.0522,
+      "step": 2214
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 4.276317445233284,
+      "learning_rate": 1.2063294094643545e-06,
+      "loss": 0.0593,
+      "step": 2215
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 3.124921604942503,
+      "learning_rate": 1.2026110951459364e-06,
+      "loss": 0.0572,
+      "step": 2216
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 3.4616042329090746,
+      "learning_rate": 1.1988977366916082e-06,
+      "loss": 0.0512,
+      "step": 2217
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 6.177271445657208,
+      "learning_rate": 1.195189338947566e-06,
+      "loss": 0.0838,
+      "step": 2218
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 4.85266593001263,
+      "learning_rate": 1.1914859067535356e-06,
+      "loss": 0.0822,
+      "step": 2219
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 4.335672007592,
+      "learning_rate": 1.18778744494276e-06,
+      "loss": 0.0664,
+      "step": 2220
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 4.171529287240308,
+      "learning_rate": 1.1840939583419986e-06,
+      "loss": 0.0746,
+      "step": 2221
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 6.495902022330293,
+      "learning_rate": 1.1804054517715125e-06,
+      "loss": 0.0886,
+      "step": 2222
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 6.3226006733851,
+      "learning_rate": 1.1767219300450694e-06,
+      "loss": 0.0625,
+      "step": 2223
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 6.92316851489635,
+      "learning_rate": 1.1730433979699235e-06,
+      "loss": 0.0959,
+      "step": 2224
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 6.580895954101497,
+      "learning_rate": 1.1693698603468296e-06,
+      "loss": 0.0991,
+      "step": 2225
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 5.079147667281248,
+      "learning_rate": 1.1657013219700108e-06,
+      "loss": 0.0764,
+      "step": 2226
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 8.490025095062611,
+      "learning_rate": 1.1620377876271767e-06,
+      "loss": 0.0988,
+      "step": 2227
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 4.705212516510039,
+      "learning_rate": 1.158379262099497e-06,
+      "loss": 0.0859,
+      "step": 2228
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 4.924318906960763,
+      "learning_rate": 1.1547257501616138e-06,
+      "loss": 0.069,
+      "step": 2229
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 6.073865969374255,
+      "learning_rate": 1.1510772565816158e-06,
+      "loss": 0.083,
+      "step": 2230
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 5.648882183192889,
+      "learning_rate": 1.1474337861210543e-06,
+      "loss": 0.0764,
+      "step": 2231
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 6.270639329248416,
+      "learning_rate": 1.1437953435349148e-06,
+      "loss": 0.0669,
+      "step": 2232
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 2.82873662808527,
+      "learning_rate": 1.1401619335716257e-06,
+      "loss": 0.0441,
+      "step": 2233
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 4.291873639982927,
+      "learning_rate": 1.1365335609730492e-06,
+      "loss": 0.0606,
+      "step": 2234
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 3.395892916999151,
+      "learning_rate": 1.1329102304744677e-06,
+      "loss": 0.0613,
+      "step": 2235
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 4.598980534640157,
+      "learning_rate": 1.1292919468045876e-06,
+      "loss": 0.0618,
+      "step": 2236
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.291531152038586,
+      "learning_rate": 1.1256787146855285e-06,
+      "loss": 0.0477,
+      "step": 2237
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 5.8526529252723405,
+      "learning_rate": 1.1220705388328186e-06,
+      "loss": 0.0726,
+      "step": 2238
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 8.84677815723228,
+      "learning_rate": 1.118467423955381e-06,
+      "loss": 0.1279,
+      "step": 2239
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.8596207539172163,
+      "learning_rate": 1.1148693747555428e-06,
+      "loss": 0.0573,
+      "step": 2240
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 4.00727467633607,
+      "learning_rate": 1.1112763959290102e-06,
+      "loss": 0.069,
+      "step": 2241
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.940740223785808,
+      "learning_rate": 1.1076884921648834e-06,
+      "loss": 0.0547,
+      "step": 2242
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 5.761853483379703,
+      "learning_rate": 1.104105668145629e-06,
+      "loss": 0.0818,
+      "step": 2243
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 3.303732248697315,
+      "learning_rate": 1.1005279285470922e-06,
+      "loss": 0.0518,
+      "step": 2244
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 4.029998279022828,
+      "learning_rate": 1.0969552780384756e-06,
+      "loss": 0.0483,
+      "step": 2245
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 4.938274106105345,
+      "learning_rate": 1.0933877212823462e-06,
+      "loss": 0.0646,
+      "step": 2246
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 6.69664350228125,
+      "learning_rate": 1.0898252629346169e-06,
+      "loss": 0.1066,
+      "step": 2247
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 3.063188111416612,
+      "learning_rate": 1.0862679076445565e-06,
+      "loss": 0.0378,
+      "step": 2248
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 6.304642987874876,
+      "learning_rate": 1.0827156600547629e-06,
+      "loss": 0.0657,
+      "step": 2249
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 4.952211105706957,
+      "learning_rate": 1.0791685248011768e-06,
+      "loss": 0.0657,
+      "step": 2250
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 4.175664519100589,
+      "learning_rate": 1.0756265065130605e-06,
+      "loss": 0.0626,
+      "step": 2251
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 3.9819610345363188,
+      "learning_rate": 1.0720896098130023e-06,
+      "loss": 0.0594,
+      "step": 2252
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 4.455864474187806,
+      "learning_rate": 1.0685578393169054e-06,
+      "loss": 0.0538,
+      "step": 2253
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 5.752748076031252,
+      "learning_rate": 1.065031199633983e-06,
+      "loss": 0.0627,
+      "step": 2254
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 5.191368379430934,
+      "learning_rate": 1.0615096953667537e-06,
+      "loss": 0.0854,
+      "step": 2255
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 5.109326233819422,
+      "learning_rate": 1.057993331111029e-06,
+      "loss": 0.0778,
+      "step": 2256
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 3.540528695943962,
+      "learning_rate": 1.054482111455919e-06,
+      "loss": 0.0568,
+      "step": 2257
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 6.952461729045498,
+      "learning_rate": 1.0509760409838128e-06,
+      "loss": 0.0576,
+      "step": 2258
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 2.9889602740861037,
+      "learning_rate": 1.0474751242703868e-06,
+      "loss": 0.0304,
+      "step": 2259
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 4.078598409393817,
+      "learning_rate": 1.0439793658845854e-06,
+      "loss": 0.0801,
+      "step": 2260
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 5.022080111667779,
+      "learning_rate": 1.0404887703886252e-06,
+      "loss": 0.0524,
+      "step": 2261
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 5.880108381786746,
+      "learning_rate": 1.0370033423379806e-06,
+      "loss": 0.0699,
+      "step": 2262
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 10.501256058874132,
+      "learning_rate": 1.0335230862813867e-06,
+      "loss": 0.1031,
+      "step": 2263
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 7.600436156657674,
+      "learning_rate": 1.0300480067608232e-06,
+      "loss": 0.0891,
+      "step": 2264
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 3.259562168585229,
+      "learning_rate": 1.0265781083115216e-06,
+      "loss": 0.0609,
+      "step": 2265
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 5.372613533178448,
+      "learning_rate": 1.0231133954619449e-06,
+      "loss": 0.0414,
+      "step": 2266
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 8.857160213490808,
+      "learning_rate": 1.0196538727337928e-06,
+      "loss": 0.0889,
+      "step": 2267
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 5.33105299181603,
+      "learning_rate": 1.0161995446419876e-06,
+      "loss": 0.0597,
+      "step": 2268
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 4.637348104170183,
+      "learning_rate": 1.012750415694676e-06,
+      "loss": 0.0565,
+      "step": 2269
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 4.570762937239995,
+      "learning_rate": 1.0093064903932175e-06,
+      "loss": 0.0697,
+      "step": 2270
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 9.548673554978985,
+      "learning_rate": 1.0058677732321826e-06,
+      "loss": 0.1064,
+      "step": 2271
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 5.879672345652387,
+      "learning_rate": 1.00243426869934e-06,
+      "loss": 0.0738,
+      "step": 2272
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 3.141325964445163,
+      "learning_rate": 9.990059812756604e-07,
+      "loss": 0.0426,
+      "step": 2273
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 3.092903366836257,
+      "learning_rate": 9.955829154353053e-07,
+      "loss": 0.0404,
+      "step": 2274
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 3.777251439113116,
+      "learning_rate": 9.921650756456164e-07,
+      "loss": 0.0558,
+      "step": 2275
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 5.0200521103265645,
+      "learning_rate": 9.887524663671243e-07,
+      "loss": 0.0742,
+      "step": 2276
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 4.159832991809378,
+      "learning_rate": 9.853450920535246e-07,
+      "loss": 0.0644,
+      "step": 2277
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 4.687185171838555,
+      "learning_rate": 9.819429571516864e-07,
+      "loss": 0.0465,
+      "step": 2278
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 3.228302604846628,
+      "learning_rate": 9.785460661016372e-07,
+      "loss": 0.0467,
+      "step": 2279
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 3.177425278413788,
+      "learning_rate": 9.751544233365634e-07,
+      "loss": 0.0442,
+      "step": 2280
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 6.184760576527297,
+      "learning_rate": 9.717680332828016e-07,
+      "loss": 0.0555,
+      "step": 2281
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 7.43785221563808,
+      "learning_rate": 9.68386900359834e-07,
+      "loss": 0.0661,
+      "step": 2282
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 3.5968605614679197,
+      "learning_rate": 9.65011028980279e-07,
+      "loss": 0.0543,
+      "step": 2283
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 7.33837420004248,
+      "learning_rate": 9.616404235498921e-07,
+      "loss": 0.0825,
+      "step": 2284
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 8.280865244542031,
+      "learning_rate": 9.582750884675523e-07,
+      "loss": 0.0665,
+      "step": 2285
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 4.934547548550383,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.0538,
+      "step": 2286
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 4.0594554597980865,
+      "learning_rate": 9.515602469081447e-07,
+      "loss": 0.0549,
+      "step": 2287
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 4.949766039775517,
+      "learning_rate": 9.482107491944265e-07,
+      "loss": 0.0577,
+      "step": 2288
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 5.9743433381237665,
+      "learning_rate": 9.448665393554413e-07,
+      "loss": 0.0697,
+      "step": 2289
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 5.524152627752683,
+      "learning_rate": 9.415276217556224e-07,
+      "loss": 0.058,
+      "step": 2290
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 9.329591381380945,
+      "learning_rate": 9.381940007524975e-07,
+      "loss": 0.1165,
+      "step": 2291
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 5.682903206575074,
+      "learning_rate": 9.348656806966777e-07,
+      "loss": 0.0737,
+      "step": 2292
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 4.777954338311628,
+      "learning_rate": 9.31542665931861e-07,
+      "loss": 0.0671,
+      "step": 2293
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 2.991300510326358,
+      "learning_rate": 9.282249607948179e-07,
+      "loss": 0.0388,
+      "step": 2294
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 6.910452992453393,
+      "learning_rate": 9.249125696153932e-07,
+      "loss": 0.1088,
+      "step": 2295
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 3.88783073756285,
+      "learning_rate": 9.216054967164916e-07,
+      "loss": 0.036,
+      "step": 2296
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 4.452800733669723,
+      "learning_rate": 9.183037464140804e-07,
+      "loss": 0.0594,
+      "step": 2297
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 5.0654841105519335,
+      "learning_rate": 9.150073230171802e-07,
+      "loss": 0.0667,
+      "step": 2298
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 5.2567250394143805,
+      "learning_rate": 9.117162308278605e-07,
+      "loss": 0.0918,
+      "step": 2299
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 5.939727117004436,
+      "learning_rate": 9.084304741412298e-07,
+      "loss": 0.0676,
+      "step": 2300
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 5.095711663360915,
+      "learning_rate": 9.051500572454375e-07,
+      "loss": 0.0739,
+      "step": 2301
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 5.532468159694068,
+      "learning_rate": 9.018749844216601e-07,
+      "loss": 0.0762,
+      "step": 2302
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 1.8890499919133887,
+      "learning_rate": 8.986052599441031e-07,
+      "loss": 0.0347,
+      "step": 2303
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 5.158757159795255,
+      "learning_rate": 8.953408880799907e-07,
+      "loss": 0.054,
+      "step": 2304
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 2.975532187175045,
+      "learning_rate": 8.920818730895625e-07,
+      "loss": 0.0368,
+      "step": 2305
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 4.024964907918647,
+      "learning_rate": 8.888282192260645e-07,
+      "loss": 0.051,
+      "step": 2306
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 4.688547286667767,
+      "learning_rate": 8.855799307357483e-07,
+      "loss": 0.0673,
+      "step": 2307
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 2.8485531903914554,
+      "learning_rate": 8.823370118578628e-07,
+      "loss": 0.0352,
+      "step": 2308
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 5.182118644609705,
+      "learning_rate": 8.790994668246499e-07,
+      "loss": 0.0605,
+      "step": 2309
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 3.111208020700417,
+      "learning_rate": 8.75867299861336e-07,
+      "loss": 0.0424,
+      "step": 2310
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 4.933187299943191,
+      "learning_rate": 8.7264051518613e-07,
+      "loss": 0.0537,
+      "step": 2311
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 5.215703872154436,
+      "learning_rate": 8.69419117010219e-07,
+      "loss": 0.06,
+      "step": 2312
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 2.902048238779602,
+      "learning_rate": 8.662031095377549e-07,
+      "loss": 0.0439,
+      "step": 2313
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 3.8537432272441365,
+      "learning_rate": 8.629924969658593e-07,
+      "loss": 0.049,
+      "step": 2314
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 5.986031643595639,
+      "learning_rate": 8.597872834846105e-07,
+      "loss": 0.0663,
+      "step": 2315
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 2.9945437702063633,
+      "learning_rate": 8.565874732770429e-07,
+      "loss": 0.0517,
+      "step": 2316
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 6.7538313481400385,
+      "learning_rate": 8.533930705191351e-07,
+      "loss": 0.0736,
+      "step": 2317
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 7.506627848583999,
+      "learning_rate": 8.502040793798133e-07,
+      "loss": 0.0826,
+      "step": 2318
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 5.098767674063163,
+      "learning_rate": 8.470205040209362e-07,
+      "loss": 0.0734,
+      "step": 2319
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 5.347742409959172,
+      "learning_rate": 8.438423485973013e-07,
+      "loss": 0.0528,
+      "step": 2320
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 3.4970528671308294,
+      "learning_rate": 8.406696172566258e-07,
+      "loss": 0.0507,
+      "step": 2321
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 3.7775468136530375,
+      "learning_rate": 8.375023141395528e-07,
+      "loss": 0.0493,
+      "step": 2322
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 3.8238051922577188,
+      "learning_rate": 8.343404433796371e-07,
+      "loss": 0.0544,
+      "step": 2323
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 5.55365447420404,
+      "learning_rate": 8.31184009103349e-07,
+      "loss": 0.0691,
+      "step": 2324
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 5.743545494105373,
+      "learning_rate": 8.28033015430057e-07,
+      "loss": 0.0741,
+      "step": 2325
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 5.21562521242854,
+      "learning_rate": 8.248874664720375e-07,
+      "loss": 0.0735,
+      "step": 2326
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 4.330352049430602,
+      "learning_rate": 8.217473663344544e-07,
+      "loss": 0.0515,
+      "step": 2327
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 5.085860113196356,
+      "learning_rate": 8.186127191153631e-07,
+      "loss": 0.0599,
+      "step": 2328
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 3.7138834664943987,
+      "learning_rate": 8.154835289057039e-07,
+      "loss": 0.0375,
+      "step": 2329
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 2.3123066435015094,
+      "learning_rate": 8.123597997892918e-07,
+      "loss": 0.0345,
+      "step": 2330
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 5.978029295412851,
+      "learning_rate": 8.092415358428174e-07,
+      "loss": 0.0601,
+      "step": 2331
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 2.95272421393182,
+      "learning_rate": 8.061287411358387e-07,
+      "loss": 0.032,
+      "step": 2332
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 2.9990763552019146,
+      "learning_rate": 8.030214197307756e-07,
+      "loss": 0.0437,
+      "step": 2333
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 6.474402321825924,
+      "learning_rate": 7.999195756829037e-07,
+      "loss": 0.0816,
+      "step": 2334
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 4.406899663197686,
+      "learning_rate": 7.968232130403525e-07,
+      "loss": 0.0524,
+      "step": 2335
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 4.3428664485424235,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.0512,
+      "step": 2336
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 5.415101124403392,
+      "learning_rate": 7.906469481279478e-07,
+      "loss": 0.0637,
+      "step": 2337
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 4.211936002554865,
+      "learning_rate": 7.875670539185626e-07,
+      "loss": 0.0461,
+      "step": 2338
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 6.175197364618274,
+      "learning_rate": 7.844926572354233e-07,
+      "loss": 0.0739,
+      "step": 2339
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 8.493426714891143,
+      "learning_rate": 7.814237620908344e-07,
+      "loss": 0.0792,
+      "step": 2340
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 5.878348650956615,
+      "learning_rate": 7.783603724899258e-07,
+      "loss": 0.0617,
+      "step": 2341
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 5.186575542590764,
+      "learning_rate": 7.753024924306374e-07,
+      "loss": 0.0881,
+      "step": 2342
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 5.570037064119628,
+      "learning_rate": 7.722501259037252e-07,
+      "loss": 0.0792,
+      "step": 2343
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 6.930574946386203,
+      "learning_rate": 7.692032768927421e-07,
+      "loss": 0.0847,
+      "step": 2344
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 3.819186545215073,
+      "learning_rate": 7.661619493740474e-07,
+      "loss": 0.0539,
+      "step": 2345
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 5.62976818612997,
+      "learning_rate": 7.631261473167878e-07,
+      "loss": 0.0897,
+      "step": 2346
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 16.653916223293404,
+      "learning_rate": 7.600958746829035e-07,
+      "loss": 0.0817,
+      "step": 2347
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 4.683483069126151,
+      "learning_rate": 7.57071135427117e-07,
+      "loss": 0.0579,
+      "step": 2348
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 12.329987381402244,
+      "learning_rate": 7.540519334969293e-07,
+      "loss": 0.0826,
+      "step": 2349
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 4.856574545412224,
+      "learning_rate": 7.51038272832616e-07,
+      "loss": 0.0507,
+      "step": 2350
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 2.8576497178846,
+      "learning_rate": 7.480301573672172e-07,
+      "loss": 0.0348,
+      "step": 2351
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 5.5906184719790595,
+      "learning_rate": 7.450275910265415e-07,
+      "loss": 0.0695,
+      "step": 2352
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 5.469455807621425,
+      "learning_rate": 7.420305777291493e-07,
+      "loss": 0.0555,
+      "step": 2353
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 5.31066856224184,
+      "learning_rate": 7.390391213863612e-07,
+      "loss": 0.0748,
+      "step": 2354
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 7.178440439384192,
+      "learning_rate": 7.36053225902239e-07,
+      "loss": 0.0805,
+      "step": 2355
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 5.037691417082863,
+      "learning_rate": 7.330728951735916e-07,
+      "loss": 0.0799,
+      "step": 2356
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 2.743741245055983,
+      "learning_rate": 7.300981330899615e-07,
+      "loss": 0.0389,
+      "step": 2357
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 19.826048255437453,
+      "learning_rate": 7.271289435336287e-07,
+      "loss": 0.1288,
+      "step": 2358
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 3.8178190163856756,
+      "learning_rate": 7.241653303795942e-07,
+      "loss": 0.0503,
+      "step": 2359
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 9.02738113425243,
+      "learning_rate": 7.212072974955891e-07,
+      "loss": 0.0544,
+      "step": 2360
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 3.7083274260957,
+      "learning_rate": 7.182548487420555e-07,
+      "loss": 0.045,
+      "step": 2361
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 4.5123533361926995,
+      "learning_rate": 7.153079879721514e-07,
+      "loss": 0.0768,
+      "step": 2362
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 5.993892620729005,
+      "learning_rate": 7.123667190317396e-07,
+      "loss": 0.0621,
+      "step": 2363
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 5.109847754054688,
+      "learning_rate": 7.094310457593867e-07,
+      "loss": 0.0724,
+      "step": 2364
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 5.59827795314608,
+      "learning_rate": 7.065009719863564e-07,
+      "loss": 0.0961,
+      "step": 2365
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 4.482254005447703,
+      "learning_rate": 7.035765015366047e-07,
+      "loss": 0.0543,
+      "step": 2366
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 3.8272575726731746,
+      "learning_rate": 7.00657638226776e-07,
+      "loss": 0.045,
+      "step": 2367
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 4.436629125246829,
+      "learning_rate": 6.977443858661937e-07,
+      "loss": 0.0665,
+      "step": 2368
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 6.549311740007576,
+      "learning_rate": 6.948367482568619e-07,
+      "loss": 0.07,
+      "step": 2369
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 9.692593461646188,
+      "learning_rate": 6.919347291934519e-07,
+      "loss": 0.1214,
+      "step": 2370
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 3.792261861358366,
+      "learning_rate": 6.890383324633121e-07,
+      "loss": 0.0607,
+      "step": 2371
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 5.0680037583491275,
+      "learning_rate": 6.86147561846443e-07,
+      "loss": 0.0658,
+      "step": 2372
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 5.913376304936262,
+      "learning_rate": 6.832624211155092e-07,
+      "loss": 0.0869,
+      "step": 2373
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 6.244044309331561,
+      "learning_rate": 6.803829140358237e-07,
+      "loss": 0.0722,
+      "step": 2374
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 5.086916943300504,
+      "learning_rate": 6.775090443653504e-07,
+      "loss": 0.0799,
+      "step": 2375
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 5.345668390460164,
+      "learning_rate": 6.746408158546947e-07,
+      "loss": 0.0677,
+      "step": 2376
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 8.290313502779792,
+      "learning_rate": 6.717782322471006e-07,
+      "loss": 0.1039,
+      "step": 2377
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 5.107185999004166,
+      "learning_rate": 6.689212972784425e-07,
+      "loss": 0.0635,
+      "step": 2378
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 3.086873454606283,
+      "learning_rate": 6.660700146772275e-07,
+      "loss": 0.0428,
+      "step": 2379
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 11.92840549590599,
+      "learning_rate": 6.632243881645805e-07,
+      "loss": 0.1082,
+      "step": 2380
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 5.059969482168196,
+      "learning_rate": 6.603844214542487e-07,
+      "loss": 0.071,
+      "step": 2381
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 11.662333258856188,
+      "learning_rate": 6.575501182525923e-07,
+      "loss": 0.0976,
+      "step": 2382
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 6.570994696046094,
+      "learning_rate": 6.547214822585802e-07,
+      "loss": 0.0641,
+      "step": 2383
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 3.5466374472622397,
+      "learning_rate": 6.518985171637826e-07,
+      "loss": 0.0397,
+      "step": 2384
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 3.9410891602661295,
+      "learning_rate": 6.490812266523716e-07,
+      "loss": 0.037,
+      "step": 2385
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 4.123785386808472,
+      "learning_rate": 6.462696144011149e-07,
+      "loss": 0.0459,
+      "step": 2386
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 5.644577490545579,
+      "learning_rate": 6.434636840793634e-07,
+      "loss": 0.0763,
+      "step": 2387
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 5.1479367299447825,
+      "learning_rate": 6.406634393490619e-07,
+      "loss": 0.0683,
+      "step": 2388
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 5.434329843880349,
+      "learning_rate": 6.378688838647257e-07,
+      "loss": 0.0613,
+      "step": 2389
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 4.656937965938182,
+      "learning_rate": 6.350800212734532e-07,
+      "loss": 0.0624,
+      "step": 2390
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 5.668233035709314,
+      "learning_rate": 6.322968552149056e-07,
+      "loss": 0.0905,
+      "step": 2391
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 4.4791260161522946,
+      "learning_rate": 6.295193893213164e-07,
+      "loss": 0.0492,
+      "step": 2392
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 6.627842080452196,
+      "learning_rate": 6.267476272174761e-07,
+      "loss": 0.0812,
+      "step": 2393
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 3.6439494986688343,
+      "learning_rate": 6.239815725207338e-07,
+      "loss": 0.0382,
+      "step": 2394
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 5.6949935073227795,
+      "learning_rate": 6.212212288409869e-07,
+      "loss": 0.0635,
+      "step": 2395
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 5.729885321952896,
+      "learning_rate": 6.184665997806832e-07,
+      "loss": 0.0707,
+      "step": 2396
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 4.4073948653375234,
+      "learning_rate": 6.157176889348082e-07,
+      "loss": 0.054,
+      "step": 2397
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 4.759887230788131,
+      "learning_rate": 6.129744998908887e-07,
+      "loss": 0.0508,
+      "step": 2398
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 4.050524496378739,
+      "learning_rate": 6.102370362289822e-07,
+      "loss": 0.0671,
+      "step": 2399
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 4.650449552818114,
+      "learning_rate": 6.075053015216759e-07,
+      "loss": 0.0557,
+      "step": 2400
+    },
+    {
+      "epoch": 3.39,
+      "eval_avg_AUC": 0.775951744669574,
+      "eval_avg_Accuracy": 0.6619280503978779,
+      "eval_avg_Accuracy-right": 0.942872049041346,
+      "eval_avg_Accuracy-wrong": 0.17204912440300205,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6599510826821385,
+      "eval_last_AUC": 0.8005162895637814,
+      "eval_last_Accuracy": 0.7417937665782494,
+      "eval_last_Accuracy-right": 0.8018129646537107,
+      "eval_last_Accuracy-wrong": 0.6371389583807141,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.679279441372939,
+      "eval_max_AUC": 0.5827436782593536,
+      "eval_max_Accuracy": 0.6359001989389921,
+      "eval_max_Accuracy-right": 0.993217686187557,
+      "eval_max_Accuracy-wrong": 0.01284967022970207,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.553431326690612,
+      "eval_min_AUC": 0.7935301302075277,
+      "eval_min_Accuracy": 0.7392241379310345,
+      "eval_min_Accuracy-right": 0.7890309116994914,
+      "eval_min_Accuracy-wrong": 0.6523766204230157,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6761883798571506,
+      "eval_prod_AUC": 0.7914367310982129,
+      "eval_prod_Accuracy": 0.7288627320954907,
+      "eval_prod_Accuracy-right": 0.7457936611451677,
+      "eval_prod_Accuracy-wrong": 0.6993404594041391,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6738920572991486,
+      "eval_runtime": 666.4466,
+      "eval_samples_per_second": 36.204,
+      "eval_steps_per_second": 1.131,
+      "eval_sum_AUC": 0.5917901696673025,
+      "eval_sum_Accuracy": 0.6347397214854111,
+      "eval_sum_Accuracy-right": 0.9940654754141124,
+      "eval_sum_Accuracy-wrong": 0.008187400500341141,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6312131580127723,
+      "step": 2400
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 3.3367433921849528,
+      "learning_rate": 6.047792993340768e-07,
+      "loss": 0.0521,
+      "step": 2401
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 3.70931709244292,
+      "learning_rate": 6.020590332238135e-07,
+      "loss": 0.0369,
+      "step": 2402
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 3.036747687843009,
+      "learning_rate": 5.993445067410286e-07,
+      "loss": 0.041,
+      "step": 2403
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 7.902684773339996,
+      "learning_rate": 5.966357234283732e-07,
+      "loss": 0.0747,
+      "step": 2404
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 4.1582334350049655,
+      "learning_rate": 5.939326868210015e-07,
+      "loss": 0.0509,
+      "step": 2405
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 4.121979859064304,
+      "learning_rate": 5.912354004465709e-07,
+      "loss": 0.0515,
+      "step": 2406
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 3.7780721868306877,
+      "learning_rate": 5.885438678252342e-07,
+      "loss": 0.048,
+      "step": 2407
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 2.8438507350798328,
+      "learning_rate": 5.858580924696317e-07,
+      "loss": 0.032,
+      "step": 2408
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 6.3866507159457,
+      "learning_rate": 5.831780778848927e-07,
+      "loss": 0.0793,
+      "step": 2409
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 3.6573839391363343,
+      "learning_rate": 5.805038275686292e-07,
+      "loss": 0.054,
+      "step": 2410
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 3.8119927393503836,
+      "learning_rate": 5.778353450109286e-07,
+      "loss": 0.0654,
+      "step": 2411
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 4.099193293246728,
+      "learning_rate": 5.751726336943509e-07,
+      "loss": 0.0688,
+      "step": 2412
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 5.959003603818399,
+      "learning_rate": 5.725156970939261e-07,
+      "loss": 0.051,
+      "step": 2413
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 5.827124909421089,
+      "learning_rate": 5.698645386771429e-07,
+      "loss": 0.0597,
+      "step": 2414
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 5.472647491882338,
+      "learning_rate": 5.67219161903958e-07,
+      "loss": 0.0864,
+      "step": 2415
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 2.5682802542473575,
+      "learning_rate": 5.645795702267731e-07,
+      "loss": 0.0396,
+      "step": 2416
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 5.3865636522123115,
+      "learning_rate": 5.619457670904466e-07,
+      "loss": 0.0791,
+      "step": 2417
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 3.9558708597667747,
+      "learning_rate": 5.593177559322776e-07,
+      "loss": 0.0474,
+      "step": 2418
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 4.921985688408169,
+      "learning_rate": 5.566955401820101e-07,
+      "loss": 0.0573,
+      "step": 2419
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 5.739244585616493,
+      "learning_rate": 5.540791232618231e-07,
+      "loss": 0.0965,
+      "step": 2420
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 4.991199811632187,
+      "learning_rate": 5.514685085863286e-07,
+      "loss": 0.0655,
+      "step": 2421
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 3.994465137409681,
+      "learning_rate": 5.488636995625635e-07,
+      "loss": 0.0504,
+      "step": 2422
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 5.07048467480604,
+      "learning_rate": 5.462646995899907e-07,
+      "loss": 0.0597,
+      "step": 2423
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 4.859309750293758,
+      "learning_rate": 5.436715120604924e-07,
+      "loss": 0.0544,
+      "step": 2424
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 2.9881282072129913,
+      "learning_rate": 5.410841403583611e-07,
+      "loss": 0.0482,
+      "step": 2425
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 5.416717966569626,
+      "learning_rate": 5.385025878603039e-07,
+      "loss": 0.0682,
+      "step": 2426
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 2.884760690012899,
+      "learning_rate": 5.359268579354304e-07,
+      "loss": 0.0301,
+      "step": 2427
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 7.4040107178669965,
+      "learning_rate": 5.333569539452543e-07,
+      "loss": 0.0764,
+      "step": 2428
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 3.1832300147258636,
+      "learning_rate": 5.307928792436812e-07,
+      "loss": 0.0401,
+      "step": 2429
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 6.1484747336849885,
+      "learning_rate": 5.282346371770147e-07,
+      "loss": 0.0714,
+      "step": 2430
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 3.3878509594032,
+      "learning_rate": 5.256822310839405e-07,
+      "loss": 0.0516,
+      "step": 2431
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 3.326631667852067,
+      "learning_rate": 5.231356642955344e-07,
+      "loss": 0.0552,
+      "step": 2432
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 5.919822383917608,
+      "learning_rate": 5.205949401352462e-07,
+      "loss": 0.0925,
+      "step": 2433
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 3.7335343564354324,
+      "learning_rate": 5.180600619189036e-07,
+      "loss": 0.0626,
+      "step": 2434
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 4.462850638321735,
+      "learning_rate": 5.155310329547025e-07,
+      "loss": 0.0624,
+      "step": 2435
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 4.178957080308394,
+      "learning_rate": 5.130078565432089e-07,
+      "loss": 0.0537,
+      "step": 2436
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 5.126851450131987,
+      "learning_rate": 5.104905359773455e-07,
+      "loss": 0.0485,
+      "step": 2437
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 2.861295387662897,
+      "learning_rate": 5.079790745423985e-07,
+      "loss": 0.0349,
+      "step": 2438
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 3.9823546154657983,
+      "learning_rate": 5.054734755160029e-07,
+      "loss": 0.0693,
+      "step": 2439
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 10.4424522490629,
+      "learning_rate": 5.029737421681446e-07,
+      "loss": 0.1049,
+      "step": 2440
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 4.754028456746745,
+      "learning_rate": 5.004798777611564e-07,
+      "loss": 0.0638,
+      "step": 2441
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 7.487306960853728,
+      "learning_rate": 4.979918855497074e-07,
+      "loss": 0.0922,
+      "step": 2442
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 7.891483816021357,
+      "learning_rate": 4.955097687808058e-07,
+      "loss": 0.0918,
+      "step": 2443
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 11.139478774097922,
+      "learning_rate": 4.930335306937923e-07,
+      "loss": 0.0716,
+      "step": 2444
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 3.6918467333252623,
+      "learning_rate": 4.905631745203348e-07,
+      "loss": 0.0484,
+      "step": 2445
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 4.878352985487895,
+      "learning_rate": 4.880987034844231e-07,
+      "loss": 0.0792,
+      "step": 2446
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 2.90002472614023,
+      "learning_rate": 4.856401208023709e-07,
+      "loss": 0.042,
+      "step": 2447
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 3.010124389659069,
+      "learning_rate": 4.831874296828004e-07,
+      "loss": 0.0334,
+      "step": 2448
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 5.766055613884313,
+      "learning_rate": 4.80740633326654e-07,
+      "loss": 0.071,
+      "step": 2449
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 9.167071352717867,
+      "learning_rate": 4.782997349271717e-07,
+      "loss": 0.0895,
+      "step": 2450
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 3.733491126964277,
+      "learning_rate": 4.758647376699033e-07,
+      "loss": 0.0495,
+      "step": 2451
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 2.6868229595360837,
+      "learning_rate": 4.7343564473269156e-07,
+      "loss": 0.0268,
+      "step": 2452
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 4.185135428267768,
+      "learning_rate": 4.710124592856796e-07,
+      "loss": 0.0541,
+      "step": 2453
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 4.954251522708119,
+      "learning_rate": 4.6859518449129393e-07,
+      "loss": 0.0461,
+      "step": 2454
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 7.620470031891302,
+      "learning_rate": 4.661838235042554e-07,
+      "loss": 0.0634,
+      "step": 2455
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 10.175645541066444,
+      "learning_rate": 4.637783794715589e-07,
+      "loss": 0.0856,
+      "step": 2456
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 12.968677064415788,
+      "learning_rate": 4.6137885553248387e-07,
+      "loss": 0.0974,
+      "step": 2457
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 4.148637212790125,
+      "learning_rate": 4.5898525481857804e-07,
+      "loss": 0.0611,
+      "step": 2458
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 5.9610400576575,
+      "learning_rate": 4.5659758045366265e-07,
+      "loss": 0.0728,
+      "step": 2459
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 3.4181745431886053,
+      "learning_rate": 4.542158355538234e-07,
+      "loss": 0.0478,
+      "step": 2460
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 2.906108772228672,
+      "learning_rate": 4.5184002322740784e-07,
+      "loss": 0.0378,
+      "step": 2461
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 3.781118869231973,
+      "learning_rate": 4.494701465750217e-07,
+      "loss": 0.0576,
+      "step": 2462
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 5.332999274021929,
+      "learning_rate": 4.4710620868952224e-07,
+      "loss": 0.0654,
+      "step": 2463
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 4.885980027608736,
+      "learning_rate": 4.4474821265601874e-07,
+      "loss": 0.0533,
+      "step": 2464
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 3.1478905950579,
+      "learning_rate": 4.423961615518618e-07,
+      "loss": 0.0408,
+      "step": 2465
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 4.658906060105995,
+      "learning_rate": 4.400500584466505e-07,
+      "loss": 0.0562,
+      "step": 2466
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 4.225306749709684,
+      "learning_rate": 4.377099064022139e-07,
+      "loss": 0.0406,
+      "step": 2467
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 3.220871343881357,
+      "learning_rate": 4.353757084726201e-07,
+      "loss": 0.0593,
+      "step": 2468
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 7.365197646977598,
+      "learning_rate": 4.3304746770416184e-07,
+      "loss": 0.0818,
+      "step": 2469
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 2.9300735164074427,
+      "learning_rate": 4.307251871353618e-07,
+      "loss": 0.0341,
+      "step": 2470
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 9.762987060589811,
+      "learning_rate": 4.2840886979696074e-07,
+      "loss": 0.0965,
+      "step": 2471
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 15.231386590879282,
+      "learning_rate": 4.2609851871192045e-07,
+      "loss": 0.0844,
+      "step": 2472
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 5.877885072354597,
+      "learning_rate": 4.237941368954124e-07,
+      "loss": 0.0662,
+      "step": 2473
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 3.1929013074594823,
+      "learning_rate": 4.214957273548209e-07,
+      "loss": 0.0468,
+      "step": 2474
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 4.5300121528286015,
+      "learning_rate": 4.1920329308973385e-07,
+      "loss": 0.0586,
+      "step": 2475
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 3.282909904219711,
+      "learning_rate": 4.1691683709194184e-07,
+      "loss": 0.0367,
+      "step": 2476
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 3.063790036305034,
+      "learning_rate": 4.146363623454341e-07,
+      "loss": 0.0383,
+      "step": 2477
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 4.883824155094312,
+      "learning_rate": 4.1236187182639386e-07,
+      "loss": 0.058,
+      "step": 2478
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 5.120138117674605,
+      "learning_rate": 4.1009336850319213e-07,
+      "loss": 0.0564,
+      "step": 2479
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 9.872069488308481,
+      "learning_rate": 4.0783085533638913e-07,
+      "loss": 0.0545,
+      "step": 2480
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 4.260845209594039,
+      "learning_rate": 4.0557433527872667e-07,
+      "loss": 0.063,
+      "step": 2481
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 6.530053407641351,
+      "learning_rate": 4.033238112751231e-07,
+      "loss": 0.0628,
+      "step": 2482
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 10.386374302555794,
+      "learning_rate": 4.0107928626267555e-07,
+      "loss": 0.1194,
+      "step": 2483
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 3.86400500433283,
+      "learning_rate": 3.9884076317064813e-07,
+      "loss": 0.0463,
+      "step": 2484
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 4.436658683861635,
+      "learning_rate": 3.9660824492047503e-07,
+      "loss": 0.052,
+      "step": 2485
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 5.2248050081532735,
+      "learning_rate": 3.9438173442575e-07,
+      "loss": 0.0499,
+      "step": 2486
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 4.559125279732787,
+      "learning_rate": 3.921612345922293e-07,
+      "loss": 0.0386,
+      "step": 2487
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 9.839602493643781,
+      "learning_rate": 3.899467483178243e-07,
+      "loss": 0.054,
+      "step": 2488
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 4.6346003111580485,
+      "learning_rate": 3.8773827849259927e-07,
+      "loss": 0.0593,
+      "step": 2489
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 4.058434488110774,
+      "learning_rate": 3.855358279987631e-07,
+      "loss": 0.0584,
+      "step": 2490
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 3.982371221579953,
+      "learning_rate": 3.833393997106727e-07,
+      "loss": 0.047,
+      "step": 2491
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 3.461951090642073,
+      "learning_rate": 3.8114899649482284e-07,
+      "loss": 0.0574,
+      "step": 2492
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 5.161895022460522,
+      "learning_rate": 3.7896462120984636e-07,
+      "loss": 0.0697,
+      "step": 2493
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 12.408145055112687,
+      "learning_rate": 3.7678627670651057e-07,
+      "loss": 0.0973,
+      "step": 2494
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 8.279259324621753,
+      "learning_rate": 3.7461396582771035e-07,
+      "loss": 0.0802,
+      "step": 2495
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 13.47364842290208,
+      "learning_rate": 3.724476914084657e-07,
+      "loss": 0.0716,
+      "step": 2496
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 4.491595580033273,
+      "learning_rate": 3.7028745627592066e-07,
+      "loss": 0.0604,
+      "step": 2497
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 6.3225119376383185,
+      "learning_rate": 3.681332632493362e-07,
+      "loss": 0.0715,
+      "step": 2498
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 7.272078567591171,
+      "learning_rate": 3.659851151400895e-07,
+      "loss": 0.0773,
+      "step": 2499
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 2.8090564876374673,
+      "learning_rate": 3.638430147516653e-07,
+      "loss": 0.0408,
+      "step": 2500
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 9.577090172604342,
+      "learning_rate": 3.6170696487965894e-07,
+      "loss": 0.0879,
+      "step": 2501
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 5.233537663334114,
+      "learning_rate": 3.595769683117689e-07,
+      "loss": 0.0533,
+      "step": 2502
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 9.89794011724972,
+      "learning_rate": 3.5745302782779157e-07,
+      "loss": 0.0777,
+      "step": 2503
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 6.4216965738627,
+      "learning_rate": 3.553351461996218e-07,
+      "loss": 0.0995,
+      "step": 2504
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 3.9426521629856106,
+      "learning_rate": 3.5322332619124657e-07,
+      "loss": 0.0515,
+      "step": 2505
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 6.5921276546651955,
+      "learning_rate": 3.511175705587433e-07,
+      "loss": 0.0724,
+      "step": 2506
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 3.5642309762182633,
+      "learning_rate": 3.4901788205027154e-07,
+      "loss": 0.0585,
+      "step": 2507
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 5.809406931359942,
+      "learning_rate": 3.469242634060771e-07,
+      "loss": 0.0585,
+      "step": 2508
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 3.4891125735343733,
+      "learning_rate": 3.448367173584799e-07,
+      "loss": 0.0549,
+      "step": 2509
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 7.537205118410471,
+      "learning_rate": 3.427552466318773e-07,
+      "loss": 0.059,
+      "step": 2510
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 3.1791154040800187,
+      "learning_rate": 3.406798539427386e-07,
+      "loss": 0.0423,
+      "step": 2511
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 3.468136564499123,
+      "learning_rate": 3.386105419995994e-07,
+      "loss": 0.0333,
+      "step": 2512
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 3.474604426684384,
+      "learning_rate": 3.3654731350305893e-07,
+      "loss": 0.0435,
+      "step": 2513
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 7.399958344801389,
+      "learning_rate": 3.344901711457782e-07,
+      "loss": 0.0988,
+      "step": 2514
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 5.439989458352775,
+      "learning_rate": 3.324391176124753e-07,
+      "loss": 0.0732,
+      "step": 2515
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 2.6978646637478465,
+      "learning_rate": 3.303941555799223e-07,
+      "loss": 0.0305,
+      "step": 2516
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 4.739477903738438,
+      "learning_rate": 3.283552877169399e-07,
+      "loss": 0.0397,
+      "step": 2517
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 3.9401533750798103,
+      "learning_rate": 3.2632251668439695e-07,
+      "loss": 0.0565,
+      "step": 2518
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 7.801485560913015,
+      "learning_rate": 3.2429584513520576e-07,
+      "loss": 0.0843,
+      "step": 2519
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 5.030913460073456,
+      "learning_rate": 3.22275275714316e-07,
+      "loss": 0.0699,
+      "step": 2520
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 4.2908901212346775,
+      "learning_rate": 3.2026081105871634e-07,
+      "loss": 0.0627,
+      "step": 2521
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 4.997168337220835,
+      "learning_rate": 3.1825245379742687e-07,
+      "loss": 0.0694,
+      "step": 2522
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 4.001937184311608,
+      "learning_rate": 3.162502065514983e-07,
+      "loss": 0.0383,
+      "step": 2523
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 8.996363914047455,
+      "learning_rate": 3.142540719340054e-07,
+      "loss": 0.0863,
+      "step": 2524
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 4.519818740839438,
+      "learning_rate": 3.12264052550047e-07,
+      "loss": 0.0549,
+      "step": 2525
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 4.250844385763064,
+      "learning_rate": 3.1028015099673957e-07,
+      "loss": 0.046,
+      "step": 2526
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 2.563861939891289,
+      "learning_rate": 3.0830236986321926e-07,
+      "loss": 0.0343,
+      "step": 2527
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 8.064669790471575,
+      "learning_rate": 3.0633071173062966e-07,
+      "loss": 0.0829,
+      "step": 2528
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 3.7825491952630697,
+      "learning_rate": 3.0436517917212694e-07,
+      "loss": 0.0507,
+      "step": 2529
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 3.8997623454351955,
+      "learning_rate": 3.024057747528697e-07,
+      "loss": 0.0539,
+      "step": 2530
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 6.097445200653237,
+      "learning_rate": 3.00452501030023e-07,
+      "loss": 0.0676,
+      "step": 2531
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 3.4416411202829904,
+      "learning_rate": 2.985053605527466e-07,
+      "loss": 0.0373,
+      "step": 2532
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 3.3183399498854684,
+      "learning_rate": 2.965643558622006e-07,
+      "loss": 0.0443,
+      "step": 2533
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 4.674537857132207,
+      "learning_rate": 2.9462948949153305e-07,
+      "loss": 0.0725,
+      "step": 2534
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 3.444782310915863,
+      "learning_rate": 2.92700763965883e-07,
+      "loss": 0.0438,
+      "step": 2535
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 5.456644812948672,
+      "learning_rate": 2.9077818180237693e-07,
+      "loss": 0.0687,
+      "step": 2536
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 3.2609835899262163,
+      "learning_rate": 2.8886174551012e-07,
+      "loss": 0.0445,
+      "step": 2537
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 7.653546611460115,
+      "learning_rate": 2.869514575901994e-07,
+      "loss": 0.0818,
+      "step": 2538
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 4.782311904189908,
+      "learning_rate": 2.850473205356774e-07,
+      "loss": 0.0695,
+      "step": 2539
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 5.0471699705890165,
+      "learning_rate": 2.831493368315902e-07,
+      "loss": 0.0428,
+      "step": 2540
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 16.342025536939868,
+      "learning_rate": 2.812575089549402e-07,
+      "loss": 0.1918,
+      "step": 2541
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 8.113735609623701,
+      "learning_rate": 2.793718393746997e-07,
+      "loss": 0.1428,
+      "step": 2542
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 4.461829366664454,
+      "learning_rate": 2.774923305518007e-07,
+      "loss": 0.0654,
+      "step": 2543
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 4.844154177765957,
+      "learning_rate": 2.756189849391389e-07,
+      "loss": 0.0727,
+      "step": 2544
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 3.74202422692227,
+      "learning_rate": 2.737518049815624e-07,
+      "loss": 0.0373,
+      "step": 2545
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 3.665281097257399,
+      "learning_rate": 2.7189079311587596e-07,
+      "loss": 0.0562,
+      "step": 2546
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 4.063645238268401,
+      "learning_rate": 2.700359517708312e-07,
+      "loss": 0.0542,
+      "step": 2547
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 3.4629215984466053,
+      "learning_rate": 2.681872833671312e-07,
+      "loss": 0.051,
+      "step": 2548
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 3.3254005912821096,
+      "learning_rate": 2.6634479031741766e-07,
+      "loss": 0.0479,
+      "step": 2549
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 8.894404998400127,
+      "learning_rate": 2.6450847502627883e-07,
+      "loss": 0.0827,
+      "step": 2550
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 4.5271982559224915,
+      "learning_rate": 2.626783398902355e-07,
+      "loss": 0.0533,
+      "step": 2551
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 5.059803244293782,
+      "learning_rate": 2.6085438729774483e-07,
+      "loss": 0.0712,
+      "step": 2552
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.853070448514815,
+      "learning_rate": 2.5903661962919713e-07,
+      "loss": 0.0429,
+      "step": 2553
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 4.152697822365192,
+      "learning_rate": 2.57225039256907e-07,
+      "loss": 0.0673,
+      "step": 2554
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.70496263462927,
+      "learning_rate": 2.5541964854511715e-07,
+      "loss": 0.04,
+      "step": 2555
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 6.765737837739038,
+      "learning_rate": 2.536204498499922e-07,
+      "loss": 0.0828,
+      "step": 2556
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 7.433603434089136,
+      "learning_rate": 2.518274455196151e-07,
+      "loss": 0.0971,
+      "step": 2557
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 4.110739394064037,
+      "learning_rate": 2.5004063789398405e-07,
+      "loss": 0.062,
+      "step": 2558
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 4.179976700926051,
+      "learning_rate": 2.4826002930501135e-07,
+      "loss": 0.0466,
+      "step": 2559
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 4.002556139772245,
+      "learning_rate": 2.464856220765172e-07,
+      "loss": 0.0504,
+      "step": 2560
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 16.684733221707486,
+      "learning_rate": 2.447174185242324e-07,
+      "loss": 0.0849,
+      "step": 2561
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 6.735054850060878,
+      "learning_rate": 2.4295542095578664e-07,
+      "loss": 0.0703,
+      "step": 2562
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 4.664541865756088,
+      "learning_rate": 2.4119963167071494e-07,
+      "loss": 0.0629,
+      "step": 2563
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 4.189436398980666,
+      "learning_rate": 2.3945005296044544e-07,
+      "loss": 0.0547,
+      "step": 2564
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 2.915969549226384,
+      "learning_rate": 2.3770668710830502e-07,
+      "loss": 0.029,
+      "step": 2565
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 3.1279783491690543,
+      "learning_rate": 2.3596953638951093e-07,
+      "loss": 0.0427,
+      "step": 2566
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 4.4893174271581895,
+      "learning_rate": 2.3423860307116853e-07,
+      "loss": 0.0517,
+      "step": 2567
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 4.865351684284176,
+      "learning_rate": 2.325138894122697e-07,
+      "loss": 0.0619,
+      "step": 2568
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 3.750933098610607,
+      "learning_rate": 2.3079539766368952e-07,
+      "loss": 0.0595,
+      "step": 2569
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 3.6420160147516225,
+      "learning_rate": 2.2908313006818171e-07,
+      "loss": 0.0466,
+      "step": 2570
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 5.063074662522242,
+      "learning_rate": 2.2737708886037823e-07,
+      "loss": 0.0667,
+      "step": 2571
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 8.535854646109012,
+      "learning_rate": 2.2567727626678527e-07,
+      "loss": 0.0788,
+      "step": 2572
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 3.8874235746990906,
+      "learning_rate": 2.239836945057794e-07,
+      "loss": 0.0424,
+      "step": 2573
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 5.134591139234094,
+      "learning_rate": 2.2229634578760705e-07,
+      "loss": 0.0559,
+      "step": 2574
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 5.590891075055715,
+      "learning_rate": 2.2061523231437733e-07,
+      "loss": 0.0708,
+      "step": 2575
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 8.571981588127402,
+      "learning_rate": 2.1894035628006517e-07,
+      "loss": 0.1155,
+      "step": 2576
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 5.871449944013923,
+      "learning_rate": 2.1727171987050156e-07,
+      "loss": 0.0583,
+      "step": 2577
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 3.5992656042066162,
+      "learning_rate": 2.156093252633784e-07,
+      "loss": 0.0798,
+      "step": 2578
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 3.870620592402866,
+      "learning_rate": 2.1395317462823807e-07,
+      "loss": 0.0393,
+      "step": 2579
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 5.907101749790395,
+      "learning_rate": 2.1230327012647713e-07,
+      "loss": 0.0716,
+      "step": 2580
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 4.435792472563452,
+      "learning_rate": 2.1065961391133705e-07,
+      "loss": 0.056,
+      "step": 2581
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 4.034549105923269,
+      "learning_rate": 2.0902220812790752e-07,
+      "loss": 0.0608,
+      "step": 2582
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 7.529461740339036,
+      "learning_rate": 2.0739105491312028e-07,
+      "loss": 0.0632,
+      "step": 2583
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 2.511784836442004,
+      "learning_rate": 2.0576615639574803e-07,
+      "loss": 0.0313,
+      "step": 2584
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 4.151319509638439,
+      "learning_rate": 2.0414751469639726e-07,
+      "loss": 0.0584,
+      "step": 2585
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 5.460010625905737,
+      "learning_rate": 2.0253513192751374e-07,
+      "loss": 0.0555,
+      "step": 2586
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 5.358638704743427,
+      "learning_rate": 2.0092901019336976e-07,
+      "loss": 0.0665,
+      "step": 2587
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 4.062216860790402,
+      "learning_rate": 1.9932915159007027e-07,
+      "loss": 0.0613,
+      "step": 2588
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 4.712105747286395,
+      "learning_rate": 1.9773555820554512e-07,
+      "loss": 0.0468,
+      "step": 2589
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 4.908125820339754,
+      "learning_rate": 1.9614823211954847e-07,
+      "loss": 0.082,
+      "step": 2590
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 4.430534523567957,
+      "learning_rate": 1.9456717540365267e-07,
+      "loss": 0.0479,
+      "step": 2591
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 5.4162513237426495,
+      "learning_rate": 1.929923901212505e-07,
+      "loss": 0.0607,
+      "step": 2592
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 3.9901284433701356,
+      "learning_rate": 1.9142387832754961e-07,
+      "loss": 0.0426,
+      "step": 2593
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 3.535726790285873,
+      "learning_rate": 1.8986164206957037e-07,
+      "loss": 0.0352,
+      "step": 2594
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 4.349532478225481,
+      "learning_rate": 1.8830568338614242e-07,
+      "loss": 0.0576,
+      "step": 2595
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 9.086163994599474,
+      "learning_rate": 1.867560043079031e-07,
+      "loss": 0.0775,
+      "step": 2596
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 5.1456318129769265,
+      "learning_rate": 1.852126068572946e-07,
+      "loss": 0.062,
+      "step": 2597
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 4.914187385091923,
+      "learning_rate": 1.836754930485607e-07,
+      "loss": 0.0721,
+      "step": 2598
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 5.1558746246377565,
+      "learning_rate": 1.8214466488774452e-07,
+      "loss": 0.0632,
+      "step": 2599
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 4.009726753470728,
+      "learning_rate": 1.8062012437268683e-07,
+      "loss": 0.0477,
+      "step": 2600
+    },
+    {
+      "epoch": 3.67,
+      "eval_avg_AUC": 0.7802949384279373,
+      "eval_avg_Accuracy": 0.6608504641909815,
+      "eval_avg_Accuracy-right": 0.9494587191861223,
+      "eval_avg_Accuracy-wrong": 0.15760745963156697,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6599857756045528,
+      "eval_last_AUC": 0.8019884262183693,
+      "eval_last_Accuracy": 0.7441561671087533,
+      "eval_last_Accuracy-right": 0.8149863049432633,
+      "eval_last_Accuracy-wrong": 0.6206504434841937,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6803915911295377,
+      "eval_max_AUC": 0.5819787457797788,
+      "eval_max_Accuracy": 0.6354442970822282,
+      "eval_max_Accuracy-right": 0.9938046171905569,
+      "eval_max_Accuracy-wrong": 0.010575392312940642,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5466072963176019,
+      "eval_min_AUC": 0.7951490552313745,
+      "eval_min_Accuracy": 0.7417523209549072,
+      "eval_min_Accuracy-right": 0.8014216773183774,
+      "eval_min_Accuracy-wrong": 0.6377075278599045,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6790361953169938,
+      "eval_prod_AUC": 0.7930242946987196,
+      "eval_prod_Accuracy": 0.7377735411140584,
+      "eval_prod_Accuracy-right": 0.7680970392591626,
+      "eval_prod_Accuracy-wrong": 0.6848987946327041,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6741797890178456,
+      "eval_runtime": 666.3302,
+      "eval_samples_per_second": 36.21,
+      "eval_steps_per_second": 1.132,
+      "eval_sum_AUC": 0.5932772120823335,
+      "eval_sum_Accuracy": 0.6345324933687002,
+      "eval_sum_Accuracy-right": 0.994326333637668,
+      "eval_sum_Accuracy-wrong": 0.007163975437798499,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6335651103494555,
+      "step": 2600
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 4.365117404414692,
+      "learning_rate": 1.7910187349302278e-07,
+      "loss": 0.0507,
+      "step": 2601
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 5.9836075505376805,
+      "learning_rate": 1.775899142301768e-07,
+      "loss": 0.077,
+      "step": 2602
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 4.946371923262528,
+      "learning_rate": 1.7608424855736494e-07,
+      "loss": 0.0544,
+      "step": 2603
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 4.477607094825463,
+      "learning_rate": 1.745848784395876e-07,
+      "loss": 0.0455,
+      "step": 2604
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 5.800920026501513,
+      "learning_rate": 1.7309180583363062e-07,
+      "loss": 0.1042,
+      "step": 2605
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 3.1634619825692365,
+      "learning_rate": 1.7160503268806084e-07,
+      "loss": 0.0367,
+      "step": 2606
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 5.422829810450787,
+      "learning_rate": 1.7012456094322337e-07,
+      "loss": 0.0605,
+      "step": 2607
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 4.758922831929125,
+      "learning_rate": 1.6865039253123938e-07,
+      "loss": 0.0604,
+      "step": 2608
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 6.483561465656901,
+      "learning_rate": 1.6718252937600376e-07,
+      "loss": 0.0772,
+      "step": 2609
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 6.0407889626016065,
+      "learning_rate": 1.657209733931836e-07,
+      "loss": 0.0846,
+      "step": 2610
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 2.4054775364338568,
+      "learning_rate": 1.6426572649021477e-07,
+      "loss": 0.0357,
+      "step": 2611
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 6.7034756721743625,
+      "learning_rate": 1.628167905662964e-07,
+      "loss": 0.0781,
+      "step": 2612
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 5.811747348126727,
+      "learning_rate": 1.6137416751239476e-07,
+      "loss": 0.0693,
+      "step": 2613
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 7.518302483580912,
+      "learning_rate": 1.599378592112366e-07,
+      "loss": 0.0857,
+      "step": 2614
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 3.7953050715469523,
+      "learning_rate": 1.5850786753730585e-07,
+      "loss": 0.049,
+      "step": 2615
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 4.363556344998636,
+      "learning_rate": 1.5708419435684463e-07,
+      "loss": 0.0483,
+      "step": 2616
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 5.229577774806158,
+      "learning_rate": 1.5566684152784784e-07,
+      "loss": 0.0768,
+      "step": 2617
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 5.159020100419698,
+      "learning_rate": 1.5425581090006248e-07,
+      "loss": 0.066,
+      "step": 2618
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 3.7322942931475076,
+      "learning_rate": 1.528511043149844e-07,
+      "loss": 0.051,
+      "step": 2619
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 5.077531111772801,
+      "learning_rate": 1.5145272360585662e-07,
+      "loss": 0.0652,
+      "step": 2620
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 5.130909089476639,
+      "learning_rate": 1.5006067059766484e-07,
+      "loss": 0.0649,
+      "step": 2621
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 4.532975590903722,
+      "learning_rate": 1.4867494710713915e-07,
+      "loss": 0.0533,
+      "step": 2622
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 12.438032873807739,
+      "learning_rate": 1.472955549427463e-07,
+      "loss": 0.0986,
+      "step": 2623
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 3.951575033000227,
+      "learning_rate": 1.4592249590469344e-07,
+      "loss": 0.0526,
+      "step": 2624
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 4.797670400530669,
+      "learning_rate": 1.4455577178491942e-07,
+      "loss": 0.0546,
+      "step": 2625
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 5.940981604998143,
+      "learning_rate": 1.4319538436709746e-07,
+      "loss": 0.0593,
+      "step": 2626
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 10.590501178465692,
+      "learning_rate": 1.4184133542663014e-07,
+      "loss": 0.0866,
+      "step": 2627
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 5.072500803125589,
+      "learning_rate": 1.404936267306495e-07,
+      "loss": 0.0595,
+      "step": 2628
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 5.489808470145056,
+      "learning_rate": 1.3915226003800964e-07,
+      "loss": 0.0699,
+      "step": 2629
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 3.683360972262747,
+      "learning_rate": 1.3781723709929085e-07,
+      "loss": 0.0391,
+      "step": 2630
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 2.488837812604504,
+      "learning_rate": 1.3648855965679497e-07,
+      "loss": 0.0344,
+      "step": 2631
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 5.788801663422032,
+      "learning_rate": 1.3516622944453884e-07,
+      "loss": 0.0714,
+      "step": 2632
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 4.253658197172255,
+      "learning_rate": 1.338502481882592e-07,
+      "loss": 0.0541,
+      "step": 2633
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 5.729043379346255,
+      "learning_rate": 1.3254061760540558e-07,
+      "loss": 0.0652,
+      "step": 2634
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 4.206820028132849,
+      "learning_rate": 1.312373394051397e-07,
+      "loss": 0.0511,
+      "step": 2635
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 5.362412659241806,
+      "learning_rate": 1.2994041528833267e-07,
+      "loss": 0.0794,
+      "step": 2636
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 4.081738631881871,
+      "learning_rate": 1.2864984694756332e-07,
+      "loss": 0.0518,
+      "step": 2637
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 6.1643980312954305,
+      "learning_rate": 1.2736563606711384e-07,
+      "loss": 0.0711,
+      "step": 2638
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 4.021631736521676,
+      "learning_rate": 1.2608778432297464e-07,
+      "loss": 0.0405,
+      "step": 2639
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 3.6138816012421286,
+      "learning_rate": 1.2481629338283118e-07,
+      "loss": 0.0453,
+      "step": 2640
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 4.256089670543158,
+      "learning_rate": 1.235511649060711e-07,
+      "loss": 0.0645,
+      "step": 2641
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 6.35609512803741,
+      "learning_rate": 1.2229240054377645e-07,
+      "loss": 0.0765,
+      "step": 2642
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 3.827924440569534,
+      "learning_rate": 1.2104000193872589e-07,
+      "loss": 0.0575,
+      "step": 2643
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 4.437964543789894,
+      "learning_rate": 1.1979397072538813e-07,
+      "loss": 0.059,
+      "step": 2644
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 3.032742989919882,
+      "learning_rate": 1.1855430852992456e-07,
+      "loss": 0.0435,
+      "step": 2645
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 5.3547910766684925,
+      "learning_rate": 1.1732101697018161e-07,
+      "loss": 0.0622,
+      "step": 2646
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 4.06441908735836,
+      "learning_rate": 1.1609409765569346e-07,
+      "loss": 0.0511,
+      "step": 2647
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 4.218184475524488,
+      "learning_rate": 1.148735521876776e-07,
+      "loss": 0.046,
+      "step": 2648
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 5.971116438902539,
+      "learning_rate": 1.136593821590326e-07,
+      "loss": 0.0648,
+      "step": 2649
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 5.580181381711454,
+      "learning_rate": 1.1245158915433818e-07,
+      "loss": 0.0648,
+      "step": 2650
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 7.975728374051076,
+      "learning_rate": 1.1125017474984956e-07,
+      "loss": 0.0696,
+      "step": 2651
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 3.6243618499143575,
+      "learning_rate": 1.1005514051349974e-07,
+      "loss": 0.0452,
+      "step": 2652
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 8.224735434986021,
+      "learning_rate": 1.088664880048923e-07,
+      "loss": 0.0661,
+      "step": 2653
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 5.1535115758955845,
+      "learning_rate": 1.0768421877530521e-07,
+      "loss": 0.0732,
+      "step": 2654
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 4.001862600085173,
+      "learning_rate": 1.065083343676826e-07,
+      "loss": 0.0359,
+      "step": 2655
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 5.717902295612631,
+      "learning_rate": 1.0533883631663966e-07,
+      "loss": 0.0685,
+      "step": 2656
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 16.531122185967455,
+      "learning_rate": 1.0417572614845383e-07,
+      "loss": 0.0853,
+      "step": 2657
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 5.390321532232918,
+      "learning_rate": 1.0301900538106813e-07,
+      "loss": 0.0705,
+      "step": 2658
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 5.44941018975706,
+      "learning_rate": 1.0186867552408441e-07,
+      "loss": 0.0549,
+      "step": 2659
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 2.463403704380458,
+      "learning_rate": 1.007247380787657e-07,
+      "loss": 0.0314,
+      "step": 2660
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 3.0038166511820554,
+      "learning_rate": 9.958719453803278e-08,
+      "loss": 0.035,
+      "step": 2661
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 6.0214516496669335,
+      "learning_rate": 9.845604638646144e-08,
+      "loss": 0.0721,
+      "step": 2662
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 6.3022224160762805,
+      "learning_rate": 9.733129510028084e-08,
+      "loss": 0.0917,
+      "step": 2663
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 4.182749616288214,
+      "learning_rate": 9.62129421473712e-08,
+      "loss": 0.0483,
+      "step": 2664
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 5.494190544976551,
+      "learning_rate": 9.510098898726339e-08,
+      "loss": 0.0566,
+      "step": 2665
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 5.8847532441312715,
+      "learning_rate": 9.399543707113601e-08,
+      "loss": 0.0583,
+      "step": 2666
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 15.886113572233047,
+      "learning_rate": 9.289628784181326e-08,
+      "loss": 0.228,
+      "step": 2667
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 3.506592541609518,
+      "learning_rate": 9.180354273376324e-08,
+      "loss": 0.0458,
+      "step": 2668
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 3.42722254821804,
+      "learning_rate": 9.07172031730974e-08,
+      "loss": 0.0468,
+      "step": 2669
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 4.883501083459229,
+      "learning_rate": 8.963727057756499e-08,
+      "loss": 0.0618,
+      "step": 2670
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 4.973749853859927,
+      "learning_rate": 8.856374635655696e-08,
+      "loss": 0.0645,
+      "step": 2671
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 9.395352522787569,
+      "learning_rate": 8.749663191109759e-08,
+      "loss": 0.0995,
+      "step": 2672
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 3.571467012537315,
+      "learning_rate": 8.643592863384898e-08,
+      "loss": 0.0458,
+      "step": 2673
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 8.901784714565835,
+      "learning_rate": 8.538163790910381e-08,
+      "loss": 0.0999,
+      "step": 2674
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 3.88655484134173,
+      "learning_rate": 8.433376111278757e-08,
+      "loss": 0.0668,
+      "step": 2675
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 17.53298470343112,
+      "learning_rate": 8.329229961245355e-08,
+      "loss": 0.1048,
+      "step": 2676
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 3.524445962449328,
+      "learning_rate": 8.225725476728452e-08,
+      "loss": 0.0418,
+      "step": 2677
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 5.741922377027235,
+      "learning_rate": 8.122862792808662e-08,
+      "loss": 0.0545,
+      "step": 2678
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 6.10358204685754,
+      "learning_rate": 8.020642043729321e-08,
+      "loss": 0.0952,
+      "step": 2679
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 3.7154080550889668,
+      "learning_rate": 7.919063362895718e-08,
+      "loss": 0.0515,
+      "step": 2680
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 7.342396520521273,
+      "learning_rate": 7.818126882875254e-08,
+      "loss": 0.081,
+      "step": 2681
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 3.6930665386003496,
+      "learning_rate": 7.717832735397335e-08,
+      "loss": 0.0484,
+      "step": 2682
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 6.748461105259171,
+      "learning_rate": 7.618181051352925e-08,
+      "loss": 0.0735,
+      "step": 2683
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 4.238116273561973,
+      "learning_rate": 7.519171960794658e-08,
+      "loss": 0.0502,
+      "step": 2684
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 3.3633067083753714,
+      "learning_rate": 7.420805592936509e-08,
+      "loss": 0.0434,
+      "step": 2685
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 3.3605425019153277,
+      "learning_rate": 7.32308207615351e-08,
+      "loss": 0.0463,
+      "step": 2686
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 4.507264220980016,
+      "learning_rate": 7.22600153798192e-08,
+      "loss": 0.0487,
+      "step": 2687
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 5.4706581340281275,
+      "learning_rate": 7.129564105118725e-08,
+      "loss": 0.0737,
+      "step": 2688
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 4.763778626034156,
+      "learning_rate": 7.03376990342175e-08,
+      "loss": 0.052,
+      "step": 2689
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 4.7052014590369575,
+      "learning_rate": 6.938619057909268e-08,
+      "loss": 0.0673,
+      "step": 2690
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 5.221796306669253,
+      "learning_rate": 6.844111692759836e-08,
+      "loss": 0.0735,
+      "step": 2691
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 3.089331807633987,
+      "learning_rate": 6.75024793131246e-08,
+      "loss": 0.0435,
+      "step": 2692
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 6.5903490417764425,
+      "learning_rate": 6.657027896065982e-08,
+      "loss": 0.082,
+      "step": 2693
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 4.170875436950441,
+      "learning_rate": 6.564451708679198e-08,
+      "loss": 0.0563,
+      "step": 2694
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 5.093385917012765,
+      "learning_rate": 6.472519489970685e-08,
+      "loss": 0.0771,
+      "step": 2695
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 6.142486137113278,
+      "learning_rate": 6.381231359918638e-08,
+      "loss": 0.077,
+      "step": 2696
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 4.692440778248937,
+      "learning_rate": 6.290587437660479e-08,
+      "loss": 0.0592,
+      "step": 2697
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 4.084290856062282,
+      "learning_rate": 6.200587841493022e-08,
+      "loss": 0.0488,
+      "step": 2698
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 5.009225880035922,
+      "learning_rate": 6.11123268887226e-08,
+      "loss": 0.0461,
+      "step": 2699
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 3.3746946823109716,
+      "learning_rate": 6.022522096412909e-08,
+      "loss": 0.047,
+      "step": 2700
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 4.850681100609125,
+      "learning_rate": 5.9344561798888033e-08,
+      "loss": 0.0693,
+      "step": 2701
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 6.786244603943823,
+      "learning_rate": 5.847035054232175e-08,
+      "loss": 0.0664,
+      "step": 2702
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 5.7398138322149,
+      "learning_rate": 5.7602588335338713e-08,
+      "loss": 0.0834,
+      "step": 2703
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 4.6039801021193805,
+      "learning_rate": 5.674127631043025e-08,
+      "loss": 0.0481,
+      "step": 2704
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 6.465994391093468,
+      "learning_rate": 5.588641559166996e-08,
+      "loss": 0.0783,
+      "step": 2705
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 5.8260629200950245,
+      "learning_rate": 5.503800729471376e-08,
+      "loss": 0.0525,
+      "step": 2706
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 2.9635420859569233,
+      "learning_rate": 5.4196052526793164e-08,
+      "loss": 0.052,
+      "step": 2707
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 4.528249786922939,
+      "learning_rate": 5.3360552386720886e-08,
+      "loss": 0.0402,
+      "step": 2708
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 5.059892409672957,
+      "learning_rate": 5.25315079648836e-08,
+      "loss": 0.0758,
+      "step": 2709
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 4.259948058377573,
+      "learning_rate": 5.1708920343243596e-08,
+      "loss": 0.0423,
+      "step": 2710
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 4.8644431571033575,
+      "learning_rate": 5.089279059533658e-08,
+      "loss": 0.0742,
+      "step": 2711
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 5.8718606999538165,
+      "learning_rate": 5.008311978627001e-08,
+      "loss": 0.0572,
+      "step": 2712
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 4.433068473446638,
+      "learning_rate": 4.9279908972723054e-08,
+      "loss": 0.039,
+      "step": 2713
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 5.4659419657683275,
+      "learning_rate": 4.848315920294222e-08,
+      "loss": 0.0513,
+      "step": 2714
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 5.8987899535181985,
+      "learning_rate": 4.769287151674407e-08,
+      "loss": 0.08,
+      "step": 2715
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 7.127138971614191,
+      "learning_rate": 4.690904694550913e-08,
+      "loss": 0.0738,
+      "step": 2716
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 3.233389890730542,
+      "learning_rate": 4.61316865121858e-08,
+      "loss": 0.0438,
+      "step": 2717
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 3.8556348396728586,
+      "learning_rate": 4.536079123128478e-08,
+      "loss": 0.0547,
+      "step": 2718
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 4.004290232924131,
+      "learning_rate": 4.459636210887908e-08,
+      "loss": 0.0536,
+      "step": 2719
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 4.759869817879123,
+      "learning_rate": 4.383840014260343e-08,
+      "loss": 0.0654,
+      "step": 2720
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 5.443064893816398,
+      "learning_rate": 4.3086906321652136e-08,
+      "loss": 0.055,
+      "step": 2721
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 3.9264889031131207,
+      "learning_rate": 4.234188162677955e-08,
+      "loss": 0.0406,
+      "step": 2722
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 4.7677172299911215,
+      "learning_rate": 4.1603327030295126e-08,
+      "loss": 0.057,
+      "step": 2723
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 4.183401935166244,
+      "learning_rate": 4.087124349606564e-08,
+      "loss": 0.0545,
+      "step": 2724
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 4.475492824778985,
+      "learning_rate": 4.014563197951238e-08,
+      "loss": 0.0688,
+      "step": 2725
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 4.074082276019099,
+      "learning_rate": 3.9426493427611177e-08,
+      "loss": 0.0668,
+      "step": 2726
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 4.075895175235993,
+      "learning_rate": 3.8713828778887965e-08,
+      "loss": 0.0633,
+      "step": 2727
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 7.323428313196594,
+      "learning_rate": 3.8007638963421523e-08,
+      "loss": 0.1143,
+      "step": 2728
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 4.666639950555486,
+      "learning_rate": 3.7307924902840184e-08,
+      "loss": 0.055,
+      "step": 2729
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 3.3057406985483584,
+      "learning_rate": 3.661468751032127e-08,
+      "loss": 0.0402,
+      "step": 2730
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 5.338687182424041,
+      "learning_rate": 3.592792769058828e-08,
+      "loss": 0.0689,
+      "step": 2731
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 4.659953080949087,
+      "learning_rate": 3.524764633991262e-08,
+      "loss": 0.0667,
+      "step": 2732
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 7.133873043564425,
+      "learning_rate": 3.4573844346109665e-08,
+      "loss": 0.0585,
+      "step": 2733
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 5.251817124178878,
+      "learning_rate": 3.3906522588539324e-08,
+      "loss": 0.0743,
+      "step": 2734
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 3.413183469232597,
+      "learning_rate": 3.32456819381044e-08,
+      "loss": 0.035,
+      "step": 2735
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 22.46390575807774,
+      "learning_rate": 3.25913232572489e-08,
+      "loss": 0.1496,
+      "step": 2736
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 3.45205502574468,
+      "learning_rate": 3.194344739995803e-08,
+      "loss": 0.0471,
+      "step": 2737
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 3.2364435333933566,
+      "learning_rate": 3.1302055211756e-08,
+      "loss": 0.0461,
+      "step": 2738
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 3.3599114586112666,
+      "learning_rate": 3.06671475297049e-08,
+      "loss": 0.0428,
+      "step": 2739
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 5.085762666188053,
+      "learning_rate": 3.00387251824058e-08,
+      "loss": 0.0536,
+      "step": 2740
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 3.337452049926245,
+      "learning_rate": 2.941678898999378e-08,
+      "loss": 0.0499,
+      "step": 2741
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 4.160175007025476,
+      "learning_rate": 2.8801339764140122e-08,
+      "loss": 0.0627,
+      "step": 2742
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 4.50844078485113,
+      "learning_rate": 2.8192378308051215e-08,
+      "loss": 0.0638,
+      "step": 2743
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 3.9894670407984756,
+      "learning_rate": 2.7589905416464114e-08,
+      "loss": 0.0512,
+      "step": 2744
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 3.480358423705562,
+      "learning_rate": 2.6993921875649308e-08,
+      "loss": 0.0504,
+      "step": 2745
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 4.871133206531938,
+      "learning_rate": 2.640442846340796e-08,
+      "loss": 0.0665,
+      "step": 2746
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 4.340947535652966,
+      "learning_rate": 2.5821425949071334e-08,
+      "loss": 0.0712,
+      "step": 2747
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 3.217242838519053,
+      "learning_rate": 2.5244915093499134e-08,
+      "loss": 0.039,
+      "step": 2748
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 3.897585374895516,
+      "learning_rate": 2.467489664908007e-08,
+      "loss": 0.0583,
+      "step": 2749
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 3.6939836302462803,
+      "learning_rate": 2.4111371359727407e-08,
+      "loss": 0.0523,
+      "step": 2750
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 3.647328661956918,
+      "learning_rate": 2.355433996088341e-08,
+      "loss": 0.0506,
+      "step": 2751
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 4.7980577674671645,
+      "learning_rate": 2.300380317951323e-08,
+      "loss": 0.0712,
+      "step": 2752
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 6.749847761085827,
+      "learning_rate": 2.2459761734106578e-08,
+      "loss": 0.075,
+      "step": 2753
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 5.427437629114175,
+      "learning_rate": 2.1922216334677172e-08,
+      "loss": 0.0742,
+      "step": 2754
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 3.412774382327384,
+      "learning_rate": 2.1391167682759394e-08,
+      "loss": 0.0379,
+      "step": 2755
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 4.816768413584903,
+      "learning_rate": 2.0866616471409974e-08,
+      "loss": 0.0478,
+      "step": 2756
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 3.226779804106507,
+      "learning_rate": 2.0348563385205745e-08,
+      "loss": 0.0411,
+      "step": 2757
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 6.678832941348568,
+      "learning_rate": 1.9837009100243666e-08,
+      "loss": 0.0778,
+      "step": 2758
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 3.843764031428717,
+      "learning_rate": 1.9331954284137476e-08,
+      "loss": 0.0584,
+      "step": 2759
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 5.031210301460672,
+      "learning_rate": 1.8833399596021042e-08,
+      "loss": 0.0677,
+      "step": 2760
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 19.872422920796183,
+      "learning_rate": 1.834134568654333e-08,
+      "loss": 0.3416,
+      "step": 2761
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 5.7817582568215045,
+      "learning_rate": 1.7855793197869565e-08,
+      "loss": 0.0556,
+      "step": 2762
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.71989567974713,
+      "learning_rate": 1.737674276368062e-08,
+      "loss": 0.0342,
+      "step": 2763
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 4.707364643101285,
+      "learning_rate": 1.6904195009171954e-08,
+      "loss": 0.0726,
+      "step": 2764
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 5.3864406529974795,
+      "learning_rate": 1.643815055105191e-08,
+      "loss": 0.0624,
+      "step": 2765
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 4.902480522048409,
+      "learning_rate": 1.5978609997542306e-08,
+      "loss": 0.0657,
+      "step": 2766
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 3.311238874296493,
+      "learning_rate": 1.5525573948375062e-08,
+      "loss": 0.0453,
+      "step": 2767
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 5.488933778634793,
+      "learning_rate": 1.5079042994796123e-08,
+      "loss": 0.0583,
+      "step": 2768
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 3.1414522870291495,
+      "learning_rate": 1.4639017719559889e-08,
+      "loss": 0.0346,
+      "step": 2769
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 2.6657232886824005,
+      "learning_rate": 1.4205498696930332e-08,
+      "loss": 0.0334,
+      "step": 2770
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 5.190526713855005,
+      "learning_rate": 1.3778486492681542e-08,
+      "loss": 0.049,
+      "step": 2771
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 4.119390692831279,
+      "learning_rate": 1.3357981664093855e-08,
+      "loss": 0.0583,
+      "step": 2772
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 5.17910798572052,
+      "learning_rate": 1.2943984759957173e-08,
+      "loss": 0.0662,
+      "step": 2773
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 2.24349658913568,
+      "learning_rate": 1.2536496320567082e-08,
+      "loss": 0.0323,
+      "step": 2774
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 3.982490646243037,
+      "learning_rate": 1.2135516877724852e-08,
+      "loss": 0.0649,
+      "step": 2775
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 3.7884392300865666,
+      "learning_rate": 1.174104695473688e-08,
+      "loss": 0.0388,
+      "step": 2776
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 2.430514187991596,
+      "learning_rate": 1.1353087066415247e-08,
+      "loss": 0.0363,
+      "step": 2777
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 4.930072308686277,
+      "learning_rate": 1.0971637719074946e-08,
+      "loss": 0.0683,
+      "step": 2778
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 7.5699049004352466,
+      "learning_rate": 1.0596699410534429e-08,
+      "loss": 0.0676,
+      "step": 2779
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 3.740555750272956,
+      "learning_rate": 1.0228272630116165e-08,
+      "loss": 0.0541,
+      "step": 2780
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 6.577073377778609,
+      "learning_rate": 9.866357858642206e-09,
+      "loss": 0.0687,
+      "step": 2781
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 3.954231066182861,
+      "learning_rate": 9.510955568438062e-09,
+      "loss": 0.0503,
+      "step": 2782
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 7.143588153023444,
+      "learning_rate": 9.16206622332827e-09,
+      "loss": 0.0856,
+      "step": 2783
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 3.6489660203026695,
+      "learning_rate": 8.819690278639159e-09,
+      "loss": 0.0432,
+      "step": 2784
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 4.584246785843547,
+      "learning_rate": 8.483828181195531e-09,
+      "loss": 0.0659,
+      "step": 2785
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 6.351347833696399,
+      "learning_rate": 8.15448036932176e-09,
+      "loss": 0.0746,
+      "step": 2786
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 4.715727980189817,
+      "learning_rate": 7.831647272839583e-09,
+      "loss": 0.0668,
+      "step": 2787
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 20.071530224217028,
+      "learning_rate": 7.515329313069752e-09,
+      "loss": 0.1259,
+      "step": 2788
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 3.385370169600328,
+      "learning_rate": 7.2055269028298294e-09,
+      "loss": 0.0419,
+      "step": 2789
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 3.9779202227727395,
+      "learning_rate": 6.902240446434172e-09,
+      "loss": 0.0376,
+      "step": 2790
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 4.260173538388129,
+      "learning_rate": 6.605470339692832e-09,
+      "loss": 0.0505,
+      "step": 2791
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 3.9293306107513613,
+      "learning_rate": 6.315216969912663e-09,
+      "loss": 0.0575,
+      "step": 2792
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 25.420981338225808,
+      "learning_rate": 6.031480715894544e-09,
+      "loss": 0.2012,
+      "step": 2793
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 3.861926702348448,
+      "learning_rate": 5.754261947935047e-09,
+      "loss": 0.0477,
+      "step": 2794
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 2.5856810719595478,
+      "learning_rate": 5.483561027823659e-09,
+      "loss": 0.0337,
+      "step": 2795
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 4.048627183915361,
+      "learning_rate": 5.219378308845558e-09,
+      "loss": 0.046,
+      "step": 2796
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 3.8679233137221756,
+      "learning_rate": 4.961714135777729e-09,
+      "loss": 0.0548,
+      "step": 2797
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 4.77110438078632,
+      "learning_rate": 4.710568844889518e-09,
+      "loss": 0.0679,
+      "step": 2798
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 8.3068344705725,
+      "learning_rate": 4.465942763944297e-09,
+      "loss": 0.0898,
+      "step": 2799
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 4.737851583525239,
+      "learning_rate": 4.22783621219669e-09,
+      "loss": 0.0495,
+      "step": 2800
+    },
+    {
+      "epoch": 3.95,
+      "eval_avg_AUC": 0.7813155380702168,
+      "eval_avg_Accuracy": 0.6606017904509284,
+      "eval_avg_Accuracy-right": 0.9510238685274554,
+      "eval_avg_Accuracy-wrong": 0.15419604275642484,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6615809012662509,
+      "eval_last_AUC": 0.8021509991205157,
+      "eval_last_Accuracy": 0.745979774535809,
+      "eval_last_Accuracy-right": 0.8172688143993739,
+      "eval_last_Accuracy-wrong": 0.6216738685467365,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.680562682837235,
+      "eval_max_AUC": 0.5775475598320932,
+      "eval_max_Accuracy": 0.6354857427055703,
+      "eval_max_Accuracy-right": 0.9938046171905569,
+      "eval_max_Accuracy-wrong": 0.010689106208778713,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5480792910172085,
+      "eval_min_AUC": 0.7956004735908635,
+      "eval_min_Accuracy": 0.7439903846153846,
+      "eval_min_Accuracy-right": 0.8054649797834876,
+      "eval_min_Accuracy-wrong": 0.6367978166931999,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6794738021897385,
+      "eval_prod_AUC": 0.7933968942149898,
+      "eval_prod_Accuracy": 0.7364472811671088,
+      "eval_prod_Accuracy-right": 0.7669883918090518,
+      "eval_prod_Accuracy-wrong": 0.6831930861951331,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.674969366033623,
+      "eval_runtime": 666.0267,
+      "eval_samples_per_second": 36.227,
+      "eval_steps_per_second": 1.132,
+      "eval_sum_AUC": 0.5927349835290605,
+      "eval_sum_Accuracy": 0.6345739389920424,
+      "eval_sum_Accuracy-right": 0.994326333637668,
+      "eval_sum_Accuracy-wrong": 0.00727768933363657,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6348703927960816,
+      "step": 2800
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 5.871431483922117,
+      "learning_rate": 3.996249500392568e-09,
+      "loss": 0.0546,
+      "step": 2801
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 4.204558322744262,
+      "learning_rate": 3.771182930769057e-09,
+      "loss": 0.0428,
+      "step": 2802
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 5.989403018583124,
+      "learning_rate": 3.5526367970539765e-09,
+      "loss": 0.0978,
+      "step": 2803
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 4.990536642039405,
+      "learning_rate": 3.3406113844658415e-09,
+      "loss": 0.0651,
+      "step": 2804
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 3.63862453235812,
+      "learning_rate": 3.1351069697127536e-09,
+      "loss": 0.0453,
+      "step": 2805
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 2.145136127794524,
+      "learning_rate": 2.9361238209935085e-09,
+      "loss": 0.0275,
+      "step": 2806
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 5.702897969135944,
+      "learning_rate": 2.743662197994268e-09,
+      "loss": 0.095,
+      "step": 2807
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 3.1651140367892068,
+      "learning_rate": 2.557722351891889e-09,
+      "loss": 0.0366,
+      "step": 2808
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 4.027090090619702,
+      "learning_rate": 2.3783045253511495e-09,
+      "loss": 0.0454,
+      "step": 2809
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 9.545458131692282,
+      "learning_rate": 2.2054089525247457e-09,
+      "loss": 0.0961,
+      "step": 2810
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 4.949782679939286,
+      "learning_rate": 2.0390358590538507e-09,
+      "loss": 0.056,
+      "step": 2811
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 3.8612423400231464,
+      "learning_rate": 1.8791854620675566e-09,
+      "loss": 0.0385,
+      "step": 2812
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 4.95691880377169,
+      "learning_rate": 1.7258579701817656e-09,
+      "loss": 0.0466,
+      "step": 2813
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 6.193749630229731,
+      "learning_rate": 1.5790535835003006e-09,
+      "loss": 0.0687,
+      "step": 2814
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 4.457731409719483,
+      "learning_rate": 1.4387724936126835e-09,
+      "loss": 0.0596,
+      "step": 2815
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 4.742758627612287,
+      "learning_rate": 1.305014883595801e-09,
+      "loss": 0.0641,
+      "step": 2816
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 4.622684857693354,
+      "learning_rate": 1.1777809280139053e-09,
+      "loss": 0.058,
+      "step": 2817
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 3.0640852717066873,
+      "learning_rate": 1.0570707929152824e-09,
+      "loss": 0.0366,
+      "step": 2818
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 4.427681383318526,
+      "learning_rate": 9.428846358355836e-10,
+      "loss": 0.0692,
+      "step": 2819
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 4.3513923439541715,
+      "learning_rate": 8.35222605796715e-10,
+      "loss": 0.055,
+      "step": 2820
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 4.043188633935201,
+      "learning_rate": 7.340848433040615e-10,
+      "loss": 0.0513,
+      "step": 2821
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 2.4369130185210954,
+      "learning_rate": 6.394714803503732e-10,
+      "loss": 0.0316,
+      "step": 2822
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 8.76829738771135,
+      "learning_rate": 5.513826404129897e-10,
+      "loss": 0.08,
+      "step": 2823
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 2.3576828593991346,
+      "learning_rate": 4.698184384532845e-10,
+      "loss": 0.0354,
+      "step": 2824
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 4.490229982502265,
+      "learning_rate": 3.9477898091944135e-10,
+      "loss": 0.0565,
+      "step": 2825
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 3.6438150613182265,
+      "learning_rate": 3.262643657425679e-10,
+      "loss": 0.0562,
+      "step": 2826
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 3.271394388196574,
+      "learning_rate": 2.642746823394715e-10,
+      "loss": 0.0528,
+      "step": 2827
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 3.5559402254344348,
+      "learning_rate": 2.088100116109937e-10,
+      "loss": 0.0549,
+      "step": 2828
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 6.288012202657132,
+      "learning_rate": 1.5987042594312051e-10,
+      "loss": 0.0694,
+      "step": 2829
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 4.832014125248516,
+      "learning_rate": 1.174559892042071e-10,
+      "loss": 0.0707,
+      "step": 2830
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 2.836268408575747,
+      "learning_rate": 8.156675674941828e-11,
+      "loss": 0.0396,
+      "step": 2831
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 3.503487721395801,
+      "learning_rate": 5.220277541573282e-11,
+      "loss": 0.0472,
+      "step": 2832
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.89445615502456,
+      "learning_rate": 2.936408352582909e-11,
+      "loss": 0.0454,
+      "step": 2833
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 4.205769433965464,
+      "learning_rate": 1.3050710885864625e-11,
+      "loss": 0.0504,
+      "step": 2834
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.5689856154562962,
+      "learning_rate": 3.2626787860312593e-12,
+      "loss": 0.0582,
+      "step": 2835
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.430146293803826,
+      "learning_rate": 0.0,
+      "loss": 0.0497,
+      "step": 2836
+    },
+    {
+      "epoch": 4.0,
+      "step": 2836,
+      "total_flos": 299313487282176.0,
+      "train_loss": 0.2304679120040915,
+      "train_runtime": 28904.4092,
+      "train_samples_per_second": 6.279,
+      "train_steps_per_second": 0.098
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2836,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 24000,
+  "total_flos": 299313487282176.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}