diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,16026 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 2280,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0013157894736842105,
+      "grad_norm": 0.9943248451862411,
+      "learning_rate": 8.771929824561404e-07,
+      "loss": 1.3563,
+      "step": 1
+    },
+    {
+      "epoch": 0.002631578947368421,
+      "grad_norm": 1.0356516381932348,
+      "learning_rate": 1.7543859649122807e-06,
+      "loss": 1.3798,
+      "step": 2
+    },
+    {
+      "epoch": 0.003947368421052632,
+      "grad_norm": 0.9909513605669136,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 1.3634,
+      "step": 3
+    },
+    {
+      "epoch": 0.005263157894736842,
+      "grad_norm": 1.008401142445123,
+      "learning_rate": 3.5087719298245615e-06,
+      "loss": 1.39,
+      "step": 4
+    },
+    {
+      "epoch": 0.006578947368421052,
+      "grad_norm": 0.9851935697509421,
+      "learning_rate": 4.3859649122807014e-06,
+      "loss": 1.366,
+      "step": 5
+    },
+    {
+      "epoch": 0.007894736842105263,
+      "grad_norm": 0.9899431502727841,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 1.3624,
+      "step": 6
+    },
+    {
+      "epoch": 0.009210526315789473,
+      "grad_norm": 0.9636771026547807,
+      "learning_rate": 6.140350877192982e-06,
+      "loss": 1.3356,
+      "step": 7
+    },
+    {
+      "epoch": 0.010526315789473684,
+      "grad_norm": 0.9342738734933147,
+      "learning_rate": 7.017543859649123e-06,
+      "loss": 1.3189,
+      "step": 8
+    },
+    {
+      "epoch": 0.011842105263157895,
+      "grad_norm": 0.9962701021709995,
+      "learning_rate": 7.894736842105263e-06,
+      "loss": 1.3505,
+      "step": 9
+    },
+    {
+      "epoch": 0.013157894736842105,
+      "grad_norm": 0.942842160351429,
+      "learning_rate": 8.771929824561403e-06,
+      "loss": 1.3352,
+      "step": 10
+    },
+    {
+      "epoch": 0.014473684210526316,
+      "grad_norm": 0.926592433083301,
+      "learning_rate": 9.649122807017545e-06,
+      "loss": 1.3057,
+      "step": 11
+    },
+    {
+      "epoch": 0.015789473684210527,
+      "grad_norm": 0.8649365529618144,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 1.2733,
+      "step": 12
+    },
+    {
+      "epoch": 0.017105263157894738,
+      "grad_norm": 0.8589185686142112,
+      "learning_rate": 1.1403508771929824e-05,
+      "loss": 1.2501,
+      "step": 13
+    },
+    {
+      "epoch": 0.018421052631578946,
+      "grad_norm": 0.7827884394796245,
+      "learning_rate": 1.2280701754385964e-05,
+      "loss": 1.2434,
+      "step": 14
+    },
+    {
+      "epoch": 0.019736842105263157,
+      "grad_norm": 0.7242909203107426,
+      "learning_rate": 1.3157894736842106e-05,
+      "loss": 1.2158,
+      "step": 15
+    },
+    {
+      "epoch": 0.021052631578947368,
+      "grad_norm": 0.6258453259636476,
+      "learning_rate": 1.4035087719298246e-05,
+      "loss": 1.174,
+      "step": 16
+    },
+    {
+      "epoch": 0.02236842105263158,
+      "grad_norm": 0.581184200984528,
+      "learning_rate": 1.4912280701754386e-05,
+      "loss": 1.1096,
+      "step": 17
+    },
+    {
+      "epoch": 0.02368421052631579,
+      "grad_norm": 0.5605953225429362,
+      "learning_rate": 1.5789473684210526e-05,
+      "loss": 1.1116,
+      "step": 18
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 0.5408018123073384,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.0664,
+      "step": 19
+    },
+    {
+      "epoch": 0.02631578947368421,
+      "grad_norm": 0.5521377988070432,
+      "learning_rate": 1.7543859649122806e-05,
+      "loss": 1.0341,
+      "step": 20
+    },
+    {
+      "epoch": 0.02763157894736842,
+      "grad_norm": 0.5881737820139201,
+      "learning_rate": 1.8421052631578947e-05,
+      "loss": 0.994,
+      "step": 21
+    },
+    {
+      "epoch": 0.02894736842105263,
+      "grad_norm": 0.635129231733668,
+      "learning_rate": 1.929824561403509e-05,
+      "loss": 0.9912,
+      "step": 22
+    },
+    {
+      "epoch": 0.030263157894736843,
+      "grad_norm": 0.6075255860392289,
+      "learning_rate": 2.0175438596491227e-05,
+      "loss": 0.9355,
+      "step": 23
+    },
+    {
+      "epoch": 0.031578947368421054,
+      "grad_norm": 0.6242744152482619,
+      "learning_rate": 2.105263157894737e-05,
+      "loss": 0.8811,
+      "step": 24
+    },
+    {
+      "epoch": 0.03289473684210526,
+      "grad_norm": 0.5870365386284072,
+      "learning_rate": 2.1929824561403507e-05,
+      "loss": 0.8533,
+      "step": 25
+    },
+    {
+      "epoch": 0.034210526315789476,
+      "grad_norm": 0.59700551817252,
+      "learning_rate": 2.280701754385965e-05,
+      "loss": 0.8276,
+      "step": 26
+    },
+    {
+      "epoch": 0.035526315789473684,
+      "grad_norm": 0.5437014014509646,
+      "learning_rate": 2.368421052631579e-05,
+      "loss": 0.8294,
+      "step": 27
+    },
+    {
+      "epoch": 0.03684210526315789,
+      "grad_norm": 0.5524675830958683,
+      "learning_rate": 2.456140350877193e-05,
+      "loss": 0.7643,
+      "step": 28
+    },
+    {
+      "epoch": 0.038157894736842106,
+      "grad_norm": 0.5314274045755922,
+      "learning_rate": 2.5438596491228074e-05,
+      "loss": 0.7262,
+      "step": 29
+    },
+    {
+      "epoch": 0.039473684210526314,
+      "grad_norm": 0.47225342909039525,
+      "learning_rate": 2.6315789473684212e-05,
+      "loss": 0.6754,
+      "step": 30
+    },
+    {
+      "epoch": 0.04078947368421053,
+      "grad_norm": 0.44700447040118774,
+      "learning_rate": 2.7192982456140354e-05,
+      "loss": 0.6517,
+      "step": 31
+    },
+    {
+      "epoch": 0.042105263157894736,
+      "grad_norm": 0.4497880712541435,
+      "learning_rate": 2.8070175438596492e-05,
+      "loss": 0.5831,
+      "step": 32
+    },
+    {
+      "epoch": 0.04342105263157895,
+      "grad_norm": 0.3556582069847081,
+      "learning_rate": 2.8947368421052634e-05,
+      "loss": 0.6016,
+      "step": 33
+    },
+    {
+      "epoch": 0.04473684210526316,
+      "grad_norm": 0.26076585757651716,
+      "learning_rate": 2.9824561403508772e-05,
+      "loss": 0.578,
+      "step": 34
+    },
+    {
+      "epoch": 0.046052631578947366,
+      "grad_norm": 0.2673773341905264,
+      "learning_rate": 3.0701754385964913e-05,
+      "loss": 0.5595,
+      "step": 35
+    },
+    {
+      "epoch": 0.04736842105263158,
+      "grad_norm": 0.2456506716030528,
+      "learning_rate": 3.157894736842105e-05,
+      "loss": 0.5533,
+      "step": 36
+    },
+    {
+      "epoch": 0.04868421052631579,
+      "grad_norm": 0.22542446425364246,
+      "learning_rate": 3.24561403508772e-05,
+      "loss": 0.5515,
+      "step": 37
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.22177522330874816,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.5391,
+      "step": 38
+    },
+    {
+      "epoch": 0.05131578947368421,
+      "grad_norm": 0.2205640272585472,
+      "learning_rate": 3.421052631578947e-05,
+      "loss": 0.5154,
+      "step": 39
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 0.1985886053937116,
+      "learning_rate": 3.508771929824561e-05,
+      "loss": 0.4966,
+      "step": 40
+    },
+    {
+      "epoch": 0.05394736842105263,
+      "grad_norm": 0.17484163584959408,
+      "learning_rate": 3.5964912280701756e-05,
+      "loss": 0.5112,
+      "step": 41
+    },
+    {
+      "epoch": 0.05526315789473684,
+      "grad_norm": 0.2050282059933558,
+      "learning_rate": 3.6842105263157895e-05,
+      "loss": 0.5032,
+      "step": 42
+    },
+    {
+      "epoch": 0.056578947368421055,
+      "grad_norm": 0.21199590152167688,
+      "learning_rate": 3.771929824561404e-05,
+      "loss": 0.4952,
+      "step": 43
+    },
+    {
+      "epoch": 0.05789473684210526,
+      "grad_norm": 0.17849681012135296,
+      "learning_rate": 3.859649122807018e-05,
+      "loss": 0.4918,
+      "step": 44
+    },
+    {
+      "epoch": 0.05921052631578947,
+      "grad_norm": 0.2019786429643155,
+      "learning_rate": 3.9473684210526316e-05,
+      "loss": 0.4886,
+      "step": 45
+    },
+    {
+      "epoch": 0.060526315789473685,
+      "grad_norm": 0.20225571639981366,
+      "learning_rate": 4.0350877192982455e-05,
+      "loss": 0.4926,
+      "step": 46
+    },
+    {
+      "epoch": 0.06184210526315789,
+      "grad_norm": 0.1777962019167991,
+      "learning_rate": 4.12280701754386e-05,
+      "loss": 0.47,
+      "step": 47
+    },
+    {
+      "epoch": 0.06315789473684211,
+      "grad_norm": 0.15246060517847743,
+      "learning_rate": 4.210526315789474e-05,
+      "loss": 0.4767,
+      "step": 48
+    },
+    {
+      "epoch": 0.06447368421052632,
+      "grad_norm": 0.14898464152134963,
+      "learning_rate": 4.298245614035088e-05,
+      "loss": 0.4713,
+      "step": 49
+    },
+    {
+      "epoch": 0.06578947368421052,
+      "grad_norm": 0.13773290443803046,
+      "learning_rate": 4.3859649122807014e-05,
+      "loss": 0.4743,
+      "step": 50
+    },
+    {
+      "epoch": 0.06710526315789474,
+      "grad_norm": 0.1408929769423494,
+      "learning_rate": 4.473684210526316e-05,
+      "loss": 0.4501,
+      "step": 51
+    },
+    {
+      "epoch": 0.06842105263157895,
+      "grad_norm": 0.1368756686186497,
+      "learning_rate": 4.56140350877193e-05,
+      "loss": 0.4526,
+      "step": 52
+    },
+    {
+      "epoch": 0.06973684210526315,
+      "grad_norm": 0.13392249182682364,
+      "learning_rate": 4.649122807017544e-05,
+      "loss": 0.4347,
+      "step": 53
+    },
+    {
+      "epoch": 0.07105263157894737,
+      "grad_norm": 0.13153597559965788,
+      "learning_rate": 4.736842105263158e-05,
+      "loss": 0.4631,
+      "step": 54
+    },
+    {
+      "epoch": 0.07236842105263158,
+      "grad_norm": 0.1345131441878225,
+      "learning_rate": 4.824561403508772e-05,
+      "loss": 0.4495,
+      "step": 55
+    },
+    {
+      "epoch": 0.07368421052631578,
+      "grad_norm": 0.13825684252520345,
+      "learning_rate": 4.912280701754386e-05,
+      "loss": 0.4393,
+      "step": 56
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 0.11938240156936133,
+      "learning_rate": 5e-05,
+      "loss": 0.4216,
+      "step": 57
+    },
+    {
+      "epoch": 0.07631578947368421,
+      "grad_norm": 0.12724516953784049,
+      "learning_rate": 5.087719298245615e-05,
+      "loss": 0.4387,
+      "step": 58
+    },
+    {
+      "epoch": 0.07763157894736843,
+      "grad_norm": 0.1202517377574683,
+      "learning_rate": 5.1754385964912286e-05,
+      "loss": 0.4375,
+      "step": 59
+    },
+    {
+      "epoch": 0.07894736842105263,
+      "grad_norm": 0.11502282485893392,
+      "learning_rate": 5.2631578947368424e-05,
+      "loss": 0.43,
+      "step": 60
+    },
+    {
+      "epoch": 0.08026315789473684,
+      "grad_norm": 0.11833243534328311,
+      "learning_rate": 5.350877192982456e-05,
+      "loss": 0.4502,
+      "step": 61
+    },
+    {
+      "epoch": 0.08157894736842106,
+      "grad_norm": 0.1162576327218189,
+      "learning_rate": 5.438596491228071e-05,
+      "loss": 0.4122,
+      "step": 62
+    },
+    {
+      "epoch": 0.08289473684210526,
+      "grad_norm": 0.11646926324993098,
+      "learning_rate": 5.526315789473685e-05,
+      "loss": 0.4281,
+      "step": 63
+    },
+    {
+      "epoch": 0.08421052631578947,
+      "grad_norm": 0.10805987411499896,
+      "learning_rate": 5.6140350877192984e-05,
+      "loss": 0.4222,
+      "step": 64
+    },
+    {
+      "epoch": 0.08552631578947369,
+      "grad_norm": 0.11176891376288024,
+      "learning_rate": 5.701754385964912e-05,
+      "loss": 0.416,
+      "step": 65
+    },
+    {
+      "epoch": 0.0868421052631579,
+      "grad_norm": 0.12524757529866518,
+      "learning_rate": 5.789473684210527e-05,
+      "loss": 0.4047,
+      "step": 66
+    },
+    {
+      "epoch": 0.0881578947368421,
+      "grad_norm": 0.11006405173241035,
+      "learning_rate": 5.877192982456141e-05,
+      "loss": 0.4266,
+      "step": 67
+    },
+    {
+      "epoch": 0.08947368421052632,
+      "grad_norm": 0.12250978018085709,
+      "learning_rate": 5.9649122807017544e-05,
+      "loss": 0.4265,
+      "step": 68
+    },
+    {
+      "epoch": 0.09078947368421053,
+      "grad_norm": 0.11723491918915026,
+      "learning_rate": 6.052631578947369e-05,
+      "loss": 0.4173,
+      "step": 69
+    },
+    {
+      "epoch": 0.09210526315789473,
+      "grad_norm": 0.11327108284284121,
+      "learning_rate": 6.140350877192983e-05,
+      "loss": 0.428,
+      "step": 70
+    },
+    {
+      "epoch": 0.09342105263157895,
+      "grad_norm": 0.11624261730248432,
+      "learning_rate": 6.228070175438597e-05,
+      "loss": 0.4181,
+      "step": 71
+    },
+    {
+      "epoch": 0.09473684210526316,
+      "grad_norm": 0.1097767429764226,
+      "learning_rate": 6.31578947368421e-05,
+      "loss": 0.4172,
+      "step": 72
+    },
+    {
+      "epoch": 0.09605263157894736,
+      "grad_norm": 0.10948304839484718,
+      "learning_rate": 6.403508771929825e-05,
+      "loss": 0.4114,
+      "step": 73
+    },
+    {
+      "epoch": 0.09736842105263158,
+      "grad_norm": 0.11381746923579758,
+      "learning_rate": 6.49122807017544e-05,
+      "loss": 0.4138,
+      "step": 74
+    },
+    {
+      "epoch": 0.09868421052631579,
+      "grad_norm": 0.11879309263584711,
+      "learning_rate": 6.578947368421054e-05,
+      "loss": 0.4234,
+      "step": 75
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11504992983557438,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.4069,
+      "step": 76
+    },
+    {
+      "epoch": 0.1013157894736842,
+      "grad_norm": 0.1289568391364873,
+      "learning_rate": 6.754385964912281e-05,
+      "loss": 0.4172,
+      "step": 77
+    },
+    {
+      "epoch": 0.10263157894736842,
+      "grad_norm": 0.1027991596519212,
+      "learning_rate": 6.842105263157895e-05,
+      "loss": 0.4064,
+      "step": 78
+    },
+    {
+      "epoch": 0.10394736842105264,
+      "grad_norm": 0.1703158129160332,
+      "learning_rate": 6.929824561403509e-05,
+      "loss": 0.4146,
+      "step": 79
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.10978980803904749,
+      "learning_rate": 7.017543859649122e-05,
+      "loss": 0.3953,
+      "step": 80
+    },
+    {
+      "epoch": 0.10657894736842105,
+      "grad_norm": 0.11715465011733911,
+      "learning_rate": 7.105263157894737e-05,
+      "loss": 0.4195,
+      "step": 81
+    },
+    {
+      "epoch": 0.10789473684210527,
+      "grad_norm": 0.12584313901221594,
+      "learning_rate": 7.192982456140351e-05,
+      "loss": 0.4095,
+      "step": 82
+    },
+    {
+      "epoch": 0.10921052631578948,
+      "grad_norm": 0.11940093441042002,
+      "learning_rate": 7.280701754385966e-05,
+      "loss": 0.3918,
+      "step": 83
+    },
+    {
+      "epoch": 0.11052631578947368,
+      "grad_norm": 0.12190899565903182,
+      "learning_rate": 7.368421052631579e-05,
+      "loss": 0.3912,
+      "step": 84
+    },
+    {
+      "epoch": 0.1118421052631579,
+      "grad_norm": 0.13982192975495122,
+      "learning_rate": 7.456140350877193e-05,
+      "loss": 0.3741,
+      "step": 85
+    },
+    {
+      "epoch": 0.11315789473684211,
+      "grad_norm": 0.11868632446195661,
+      "learning_rate": 7.543859649122808e-05,
+      "loss": 0.4007,
+      "step": 86
+    },
+    {
+      "epoch": 0.11447368421052631,
+      "grad_norm": 0.12861533906094824,
+      "learning_rate": 7.631578947368422e-05,
+      "loss": 0.4054,
+      "step": 87
+    },
+    {
+      "epoch": 0.11578947368421053,
+      "grad_norm": 0.11712395700046184,
+      "learning_rate": 7.719298245614036e-05,
+      "loss": 0.3942,
+      "step": 88
+    },
+    {
+      "epoch": 0.11710526315789474,
+      "grad_norm": 0.11911543233284826,
+      "learning_rate": 7.807017543859649e-05,
+      "loss": 0.3992,
+      "step": 89
+    },
+    {
+      "epoch": 0.11842105263157894,
+      "grad_norm": 0.12597614512349814,
+      "learning_rate": 7.894736842105263e-05,
+      "loss": 0.3868,
+      "step": 90
+    },
+    {
+      "epoch": 0.11973684210526316,
+      "grad_norm": 0.11923027766958509,
+      "learning_rate": 7.982456140350878e-05,
+      "loss": 0.3994,
+      "step": 91
+    },
+    {
+      "epoch": 0.12105263157894737,
+      "grad_norm": 0.11704509885009143,
+      "learning_rate": 8.070175438596491e-05,
+      "loss": 0.3813,
+      "step": 92
+    },
+    {
+      "epoch": 0.12236842105263158,
+      "grad_norm": 0.1415178300748815,
+      "learning_rate": 8.157894736842105e-05,
+      "loss": 0.3995,
+      "step": 93
+    },
+    {
+      "epoch": 0.12368421052631579,
+      "grad_norm": 0.12356686708530311,
+      "learning_rate": 8.24561403508772e-05,
+      "loss": 0.385,
+      "step": 94
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.12008298974430553,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 0.3849,
+      "step": 95
+    },
+    {
+      "epoch": 0.12631578947368421,
+      "grad_norm": 0.11678986714389192,
+      "learning_rate": 8.421052631578948e-05,
+      "loss": 0.3763,
+      "step": 96
+    },
+    {
+      "epoch": 0.12763157894736843,
+      "grad_norm": 0.130987416202565,
+      "learning_rate": 8.508771929824562e-05,
+      "loss": 0.3795,
+      "step": 97
+    },
+    {
+      "epoch": 0.12894736842105264,
+      "grad_norm": 0.13585638455268312,
+      "learning_rate": 8.596491228070177e-05,
+      "loss": 0.3714,
+      "step": 98
+    },
+    {
+      "epoch": 0.13026315789473683,
+      "grad_norm": 0.1343330070878276,
+      "learning_rate": 8.68421052631579e-05,
+      "loss": 0.3686,
+      "step": 99
+    },
+    {
+      "epoch": 0.13157894736842105,
+      "grad_norm": 0.1344794137365553,
+      "learning_rate": 8.771929824561403e-05,
+      "loss": 0.3963,
+      "step": 100
+    },
+    {
+      "epoch": 0.13289473684210526,
+      "grad_norm": 0.13882732763835956,
+      "learning_rate": 8.859649122807017e-05,
+      "loss": 0.3759,
+      "step": 101
+    },
+    {
+      "epoch": 0.13421052631578947,
+      "grad_norm": 0.13720598333651746,
+      "learning_rate": 8.947368421052632e-05,
+      "loss": 0.3821,
+      "step": 102
+    },
+    {
+      "epoch": 0.1355263157894737,
+      "grad_norm": 0.14777326263342197,
+      "learning_rate": 9.035087719298246e-05,
+      "loss": 0.3667,
+      "step": 103
+    },
+    {
+      "epoch": 0.1368421052631579,
+      "grad_norm": 0.13337880990510198,
+      "learning_rate": 9.12280701754386e-05,
+      "loss": 0.3978,
+      "step": 104
+    },
+    {
+      "epoch": 0.13815789473684212,
+      "grad_norm": 0.14435581876340933,
+      "learning_rate": 9.210526315789474e-05,
+      "loss": 0.3751,
+      "step": 105
+    },
+    {
+      "epoch": 0.1394736842105263,
+      "grad_norm": 0.1307662742116557,
+      "learning_rate": 9.298245614035089e-05,
+      "loss": 0.3898,
+      "step": 106
+    },
+    {
+      "epoch": 0.14078947368421052,
+      "grad_norm": 0.13020669111992744,
+      "learning_rate": 9.385964912280703e-05,
+      "loss": 0.3904,
+      "step": 107
+    },
+    {
+      "epoch": 0.14210526315789473,
+      "grad_norm": 0.14245838078165218,
+      "learning_rate": 9.473684210526316e-05,
+      "loss": 0.3814,
+      "step": 108
+    },
+    {
+      "epoch": 0.14342105263157895,
+      "grad_norm": 0.13114307459864877,
+      "learning_rate": 9.56140350877193e-05,
+      "loss": 0.374,
+      "step": 109
+    },
+    {
+      "epoch": 0.14473684210526316,
+      "grad_norm": 0.12257270456238782,
+      "learning_rate": 9.649122807017544e-05,
+      "loss": 0.3835,
+      "step": 110
+    },
+    {
+      "epoch": 0.14605263157894738,
+      "grad_norm": 0.1377457018530249,
+      "learning_rate": 9.736842105263158e-05,
+      "loss": 0.3772,
+      "step": 111
+    },
+    {
+      "epoch": 0.14736842105263157,
+      "grad_norm": 0.13906509029436523,
+      "learning_rate": 9.824561403508771e-05,
+      "loss": 0.3777,
+      "step": 112
+    },
+    {
+      "epoch": 0.14868421052631578,
+      "grad_norm": 0.13702341765645296,
+      "learning_rate": 9.912280701754386e-05,
+      "loss": 0.3778,
+      "step": 113
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.131545294973937,
+      "learning_rate": 0.0001,
+      "loss": 0.3676,
+      "step": 114
+    },
+    {
+      "epoch": 0.1513157894736842,
+      "grad_norm": 0.144839378587314,
+      "learning_rate": 0.00010087719298245615,
+      "loss": 0.3656,
+      "step": 115
+    },
+    {
+      "epoch": 0.15263157894736842,
+      "grad_norm": 0.13801404126106712,
+      "learning_rate": 0.0001017543859649123,
+      "loss": 0.3698,
+      "step": 116
+    },
+    {
+      "epoch": 0.15394736842105264,
+      "grad_norm": 0.14250664033740984,
+      "learning_rate": 0.00010263157894736844,
+      "loss": 0.3556,
+      "step": 117
+    },
+    {
+      "epoch": 0.15526315789473685,
+      "grad_norm": 0.14495828716442766,
+      "learning_rate": 0.00010350877192982457,
+      "loss": 0.3874,
+      "step": 118
+    },
+    {
+      "epoch": 0.15657894736842104,
+      "grad_norm": 0.1458656065808899,
+      "learning_rate": 0.0001043859649122807,
+      "loss": 0.3707,
+      "step": 119
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.15169186829219217,
+      "learning_rate": 0.00010526315789473685,
+      "loss": 0.3807,
+      "step": 120
+    },
+    {
+      "epoch": 0.15921052631578947,
+      "grad_norm": 0.1384837433981043,
+      "learning_rate": 0.00010614035087719298,
+      "loss": 0.3606,
+      "step": 121
+    },
+    {
+      "epoch": 0.16052631578947368,
+      "grad_norm": 0.1397696554751311,
+      "learning_rate": 0.00010701754385964912,
+      "loss": 0.3703,
+      "step": 122
+    },
+    {
+      "epoch": 0.1618421052631579,
+      "grad_norm": 0.14890695623711342,
+      "learning_rate": 0.00010789473684210527,
+      "loss": 0.3613,
+      "step": 123
+    },
+    {
+      "epoch": 0.1631578947368421,
+      "grad_norm": 0.15703056192380713,
+      "learning_rate": 0.00010877192982456141,
+      "loss": 0.3504,
+      "step": 124
+    },
+    {
+      "epoch": 0.16447368421052633,
+      "grad_norm": 0.1392335253666652,
+      "learning_rate": 0.00010964912280701756,
+      "loss": 0.3708,
+      "step": 125
+    },
+    {
+      "epoch": 0.16578947368421051,
+      "grad_norm": 0.14687002166712368,
+      "learning_rate": 0.0001105263157894737,
+      "loss": 0.3698,
+      "step": 126
+    },
+    {
+      "epoch": 0.16710526315789473,
+      "grad_norm": 0.15870994148978584,
+      "learning_rate": 0.00011140350877192982,
+      "loss": 0.3779,
+      "step": 127
+    },
+    {
+      "epoch": 0.16842105263157894,
+      "grad_norm": 0.14754629529682614,
+      "learning_rate": 0.00011228070175438597,
+      "loss": 0.3623,
+      "step": 128
+    },
+    {
+      "epoch": 0.16973684210526316,
+      "grad_norm": 0.15124044879001075,
+      "learning_rate": 0.00011315789473684211,
+      "loss": 0.3608,
+      "step": 129
+    },
+    {
+      "epoch": 0.17105263157894737,
+      "grad_norm": 0.14959917991128074,
+      "learning_rate": 0.00011403508771929824,
+      "loss": 0.3642,
+      "step": 130
+    },
+    {
+      "epoch": 0.1723684210526316,
+      "grad_norm": 0.1571767924920792,
+      "learning_rate": 0.00011491228070175439,
+      "loss": 0.3631,
+      "step": 131
+    },
+    {
+      "epoch": 0.1736842105263158,
+      "grad_norm": 0.14490254678965067,
+      "learning_rate": 0.00011578947368421053,
+      "loss": 0.3559,
+      "step": 132
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 0.144229202921289,
+      "learning_rate": 0.00011666666666666668,
+      "loss": 0.3546,
+      "step": 133
+    },
+    {
+      "epoch": 0.1763157894736842,
+      "grad_norm": 0.1600537686886715,
+      "learning_rate": 0.00011754385964912282,
+      "loss": 0.3455,
+      "step": 134
+    },
+    {
+      "epoch": 0.17763157894736842,
+      "grad_norm": 0.1313150257696866,
+      "learning_rate": 0.00011842105263157894,
+      "loss": 0.3552,
+      "step": 135
+    },
+    {
+      "epoch": 0.17894736842105263,
+      "grad_norm": 0.13732726897904224,
+      "learning_rate": 0.00011929824561403509,
+      "loss": 0.3497,
+      "step": 136
+    },
+    {
+      "epoch": 0.18026315789473685,
+      "grad_norm": 0.1463571482716898,
+      "learning_rate": 0.00012017543859649123,
+      "loss": 0.3519,
+      "step": 137
+    },
+    {
+      "epoch": 0.18157894736842106,
+      "grad_norm": 0.1424850899462053,
+      "learning_rate": 0.00012105263157894738,
+      "loss": 0.3475,
+      "step": 138
+    },
+    {
+      "epoch": 0.18289473684210528,
+      "grad_norm": 0.15624853614650133,
+      "learning_rate": 0.00012192982456140352,
+      "loss": 0.3541,
+      "step": 139
+    },
+    {
+      "epoch": 0.18421052631578946,
+      "grad_norm": 0.15391671545839364,
+      "learning_rate": 0.00012280701754385965,
+      "loss": 0.347,
+      "step": 140
+    },
+    {
+      "epoch": 0.18552631578947368,
+      "grad_norm": 0.1476455568748672,
+      "learning_rate": 0.0001236842105263158,
+      "loss": 0.3641,
+      "step": 141
+    },
+    {
+      "epoch": 0.1868421052631579,
+      "grad_norm": 0.15837728677358026,
+      "learning_rate": 0.00012456140350877194,
+      "loss": 0.3542,
+      "step": 142
+    },
+    {
+      "epoch": 0.1881578947368421,
+      "grad_norm": 0.1387983783021263,
+      "learning_rate": 0.00012543859649122806,
+      "loss": 0.3425,
+      "step": 143
+    },
+    {
+      "epoch": 0.18947368421052632,
+      "grad_norm": 0.15375599222156547,
+      "learning_rate": 0.0001263157894736842,
+      "loss": 0.353,
+      "step": 144
+    },
+    {
+      "epoch": 0.19078947368421054,
+      "grad_norm": 0.14286343766479473,
+      "learning_rate": 0.00012719298245614035,
+      "loss": 0.3425,
+      "step": 145
+    },
+    {
+      "epoch": 0.19210526315789472,
+      "grad_norm": 0.15500148844719516,
+      "learning_rate": 0.0001280701754385965,
+      "loss": 0.3536,
+      "step": 146
+    },
+    {
+      "epoch": 0.19342105263157894,
+      "grad_norm": 0.1527274581882464,
+      "learning_rate": 0.00012894736842105264,
+      "loss": 0.3529,
+      "step": 147
+    },
+    {
+      "epoch": 0.19473684210526315,
+      "grad_norm": 0.13911203035920786,
+      "learning_rate": 0.0001298245614035088,
+      "loss": 0.3582,
+      "step": 148
+    },
+    {
+      "epoch": 0.19605263157894737,
+      "grad_norm": 0.1603641558907569,
+      "learning_rate": 0.00013070175438596493,
+      "loss": 0.3379,
+      "step": 149
+    },
+    {
+      "epoch": 0.19736842105263158,
+      "grad_norm": 0.15724848532335062,
+      "learning_rate": 0.00013157894736842108,
+      "loss": 0.3447,
+      "step": 150
+    },
+    {
+      "epoch": 0.1986842105263158,
+      "grad_norm": 0.15104286411263596,
+      "learning_rate": 0.0001324561403508772,
+      "loss": 0.3592,
+      "step": 151
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14256111417333361,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.3418,
+      "step": 152
+    },
+    {
+      "epoch": 0.2013157894736842,
+      "grad_norm": 0.14616033132964826,
+      "learning_rate": 0.00013421052631578948,
+      "loss": 0.365,
+      "step": 153
+    },
+    {
+      "epoch": 0.2026315789473684,
+      "grad_norm": 0.13617570529003223,
+      "learning_rate": 0.00013508771929824563,
+      "loss": 0.3687,
+      "step": 154
+    },
+    {
+      "epoch": 0.20394736842105263,
+      "grad_norm": 0.15392252191058636,
+      "learning_rate": 0.00013596491228070177,
+      "loss": 0.3481,
+      "step": 155
+    },
+    {
+      "epoch": 0.20526315789473684,
+      "grad_norm": 0.14731210087429397,
+      "learning_rate": 0.0001368421052631579,
+      "loss": 0.3578,
+      "step": 156
+    },
+    {
+      "epoch": 0.20657894736842106,
+      "grad_norm": 0.15561784274185514,
+      "learning_rate": 0.00013771929824561404,
+      "loss": 0.3592,
+      "step": 157
+    },
+    {
+      "epoch": 0.20789473684210527,
+      "grad_norm": 0.15066657618186907,
+      "learning_rate": 0.00013859649122807018,
+      "loss": 0.3467,
+      "step": 158
+    },
+    {
+      "epoch": 0.20921052631578949,
+      "grad_norm": 0.1420611399655039,
+      "learning_rate": 0.0001394736842105263,
+      "loss": 0.3616,
+      "step": 159
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.1488250715622558,
+      "learning_rate": 0.00014035087719298245,
+      "loss": 0.339,
+      "step": 160
+    },
+    {
+      "epoch": 0.2118421052631579,
+      "grad_norm": 0.1462212684656958,
+      "learning_rate": 0.0001412280701754386,
+      "loss": 0.3414,
+      "step": 161
+    },
+    {
+      "epoch": 0.2131578947368421,
+      "grad_norm": 0.13807950981607273,
+      "learning_rate": 0.00014210526315789474,
+      "loss": 0.3342,
+      "step": 162
+    },
+    {
+      "epoch": 0.21447368421052632,
+      "grad_norm": 0.15320658283775204,
+      "learning_rate": 0.00014298245614035088,
+      "loss": 0.3448,
+      "step": 163
+    },
+    {
+      "epoch": 0.21578947368421053,
+      "grad_norm": 0.1465532353318612,
+      "learning_rate": 0.00014385964912280703,
+      "loss": 0.3472,
+      "step": 164
+    },
+    {
+      "epoch": 0.21710526315789475,
+      "grad_norm": 0.14706679089630642,
+      "learning_rate": 0.00014473684210526317,
+      "loss": 0.3437,
+      "step": 165
+    },
+    {
+      "epoch": 0.21842105263157896,
+      "grad_norm": 0.15624893484875924,
+      "learning_rate": 0.00014561403508771932,
+      "loss": 0.3417,
+      "step": 166
+    },
+    {
+      "epoch": 0.21973684210526315,
+      "grad_norm": 0.14729900208480234,
+      "learning_rate": 0.00014649122807017543,
+      "loss": 0.3651,
+      "step": 167
+    },
+    {
+      "epoch": 0.22105263157894736,
+      "grad_norm": 0.13340629262561993,
+      "learning_rate": 0.00014736842105263158,
+      "loss": 0.3452,
+      "step": 168
+    },
+    {
+      "epoch": 0.22236842105263158,
+      "grad_norm": 0.1427638316007593,
+      "learning_rate": 0.00014824561403508772,
+      "loss": 0.3368,
+      "step": 169
+    },
+    {
+      "epoch": 0.2236842105263158,
+      "grad_norm": 0.1394836492366252,
+      "learning_rate": 0.00014912280701754387,
+      "loss": 0.3366,
+      "step": 170
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 0.12982465074656047,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.3266,
+      "step": 171
+    },
+    {
+      "epoch": 0.22631578947368422,
+      "grad_norm": 0.1489587612015448,
+      "learning_rate": 0.00015087719298245616,
+      "loss": 0.3342,
+      "step": 172
+    },
+    {
+      "epoch": 0.22763157894736843,
+      "grad_norm": 0.15658289578441603,
+      "learning_rate": 0.0001517543859649123,
+      "loss": 0.3461,
+      "step": 173
+    },
+    {
+      "epoch": 0.22894736842105262,
+      "grad_norm": 0.13299000630905108,
+      "learning_rate": 0.00015263157894736845,
+      "loss": 0.3343,
+      "step": 174
+    },
+    {
+      "epoch": 0.23026315789473684,
+      "grad_norm": 0.1501297297839127,
+      "learning_rate": 0.00015350877192982457,
+      "loss": 0.3319,
+      "step": 175
+    },
+    {
+      "epoch": 0.23157894736842105,
+      "grad_norm": 0.1447458092769113,
+      "learning_rate": 0.0001543859649122807,
+      "loss": 0.3538,
+      "step": 176
+    },
+    {
+      "epoch": 0.23289473684210527,
+      "grad_norm": 0.12642159107807327,
+      "learning_rate": 0.00015526315789473686,
+      "loss": 0.3507,
+      "step": 177
+    },
+    {
+      "epoch": 0.23421052631578948,
+      "grad_norm": 0.1307467882888706,
+      "learning_rate": 0.00015614035087719297,
+      "loss": 0.3352,
+      "step": 178
+    },
+    {
+      "epoch": 0.2355263157894737,
+      "grad_norm": 0.14575680367060406,
+      "learning_rate": 0.00015701754385964912,
+      "loss": 0.3415,
+      "step": 179
+    },
+    {
+      "epoch": 0.23684210526315788,
+      "grad_norm": 0.13772178353175255,
+      "learning_rate": 0.00015789473684210527,
+      "loss": 0.3517,
+      "step": 180
+    },
+    {
+      "epoch": 0.2381578947368421,
+      "grad_norm": 0.13177565526904855,
+      "learning_rate": 0.0001587719298245614,
+      "loss": 0.326,
+      "step": 181
+    },
+    {
+      "epoch": 0.2394736842105263,
+      "grad_norm": 0.13895296380351296,
+      "learning_rate": 0.00015964912280701756,
+      "loss": 0.3418,
+      "step": 182
+    },
+    {
+      "epoch": 0.24078947368421053,
+      "grad_norm": 0.13234030291334134,
+      "learning_rate": 0.0001605263157894737,
+      "loss": 0.3492,
+      "step": 183
+    },
+    {
+      "epoch": 0.24210526315789474,
+      "grad_norm": 0.1343103364584722,
+      "learning_rate": 0.00016140350877192982,
+      "loss": 0.334,
+      "step": 184
+    },
+    {
+      "epoch": 0.24342105263157895,
+      "grad_norm": 0.13330977836549726,
+      "learning_rate": 0.00016228070175438596,
+      "loss": 0.342,
+      "step": 185
+    },
+    {
+      "epoch": 0.24473684210526317,
+      "grad_norm": 0.12933054253662082,
+      "learning_rate": 0.0001631578947368421,
+      "loss": 0.3476,
+      "step": 186
+    },
+    {
+      "epoch": 0.24605263157894736,
+      "grad_norm": 0.12637139899067856,
+      "learning_rate": 0.00016403508771929825,
+      "loss": 0.3317,
+      "step": 187
+    },
+    {
+      "epoch": 0.24736842105263157,
+      "grad_norm": 0.12268533782351379,
+      "learning_rate": 0.0001649122807017544,
+      "loss": 0.3369,
+      "step": 188
+    },
+    {
+      "epoch": 0.24868421052631579,
+      "grad_norm": 0.1269784203784765,
+      "learning_rate": 0.00016578947368421054,
+      "loss": 0.3555,
+      "step": 189
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13455995822795366,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 0.3352,
+      "step": 190
+    },
+    {
+      "epoch": 0.2513157894736842,
+      "grad_norm": 0.13042417788105207,
+      "learning_rate": 0.00016754385964912283,
+      "loss": 0.3425,
+      "step": 191
+    },
+    {
+      "epoch": 0.25263157894736843,
+      "grad_norm": 0.13259270895700348,
+      "learning_rate": 0.00016842105263157895,
+      "loss": 0.3388,
+      "step": 192
+    },
+    {
+      "epoch": 0.25394736842105264,
+      "grad_norm": 0.13229060651188568,
+      "learning_rate": 0.0001692982456140351,
+      "loss": 0.3444,
+      "step": 193
+    },
+    {
+      "epoch": 0.25526315789473686,
+      "grad_norm": 0.1409002440562877,
+      "learning_rate": 0.00017017543859649124,
+      "loss": 0.326,
+      "step": 194
+    },
+    {
+      "epoch": 0.2565789473684211,
+      "grad_norm": 0.13331986255382416,
+      "learning_rate": 0.00017105263157894739,
+      "loss": 0.3333,
+      "step": 195
+    },
+    {
+      "epoch": 0.2578947368421053,
+      "grad_norm": 0.13211357309903607,
+      "learning_rate": 0.00017192982456140353,
+      "loss": 0.3303,
+      "step": 196
+    },
+    {
+      "epoch": 0.25921052631578945,
+      "grad_norm": 0.13392529140137704,
+      "learning_rate": 0.00017280701754385965,
+      "loss": 0.3266,
+      "step": 197
+    },
+    {
+      "epoch": 0.26052631578947366,
+      "grad_norm": 0.135716179073066,
+      "learning_rate": 0.0001736842105263158,
+      "loss": 0.3372,
+      "step": 198
+    },
+    {
+      "epoch": 0.2618421052631579,
+      "grad_norm": 0.1325252534873712,
+      "learning_rate": 0.00017456140350877194,
+      "loss": 0.3377,
+      "step": 199
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.13891964420010164,
+      "learning_rate": 0.00017543859649122806,
+      "loss": 0.3359,
+      "step": 200
+    },
+    {
+      "epoch": 0.2644736842105263,
+      "grad_norm": 0.14101443449526624,
+      "learning_rate": 0.0001763157894736842,
+      "loss": 0.3254,
+      "step": 201
+    },
+    {
+      "epoch": 0.2657894736842105,
+      "grad_norm": 0.12637174218931757,
+      "learning_rate": 0.00017719298245614035,
+      "loss": 0.3372,
+      "step": 202
+    },
+    {
+      "epoch": 0.26710526315789473,
+      "grad_norm": 0.13174772610887958,
+      "learning_rate": 0.0001780701754385965,
+      "loss": 0.3365,
+      "step": 203
+    },
+    {
+      "epoch": 0.26842105263157895,
+      "grad_norm": 0.13681575981535574,
+      "learning_rate": 0.00017894736842105264,
+      "loss": 0.3356,
+      "step": 204
+    },
+    {
+      "epoch": 0.26973684210526316,
+      "grad_norm": 0.1328471909403007,
+      "learning_rate": 0.00017982456140350878,
+      "loss": 0.33,
+      "step": 205
+    },
+    {
+      "epoch": 0.2710526315789474,
+      "grad_norm": 0.1407266391635488,
+      "learning_rate": 0.00018070175438596493,
+      "loss": 0.3391,
+      "step": 206
+    },
+    {
+      "epoch": 0.2723684210526316,
+      "grad_norm": 0.1348033610448726,
+      "learning_rate": 0.00018157894736842107,
+      "loss": 0.3239,
+      "step": 207
+    },
+    {
+      "epoch": 0.2736842105263158,
+      "grad_norm": 0.1265119642350682,
+      "learning_rate": 0.0001824561403508772,
+      "loss": 0.3345,
+      "step": 208
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 0.12645375412506446,
+      "learning_rate": 0.00018333333333333334,
+      "loss": 0.3241,
+      "step": 209
+    },
+    {
+      "epoch": 0.27631578947368424,
+      "grad_norm": 0.13126982713178578,
+      "learning_rate": 0.00018421052631578948,
+      "loss": 0.3367,
+      "step": 210
+    },
+    {
+      "epoch": 0.2776315789473684,
+      "grad_norm": 0.1243720842002383,
+      "learning_rate": 0.00018508771929824563,
+      "loss": 0.329,
+      "step": 211
+    },
+    {
+      "epoch": 0.2789473684210526,
+      "grad_norm": 0.14402206242169863,
+      "learning_rate": 0.00018596491228070177,
+      "loss": 0.3321,
+      "step": 212
+    },
+    {
+      "epoch": 0.2802631578947368,
+      "grad_norm": 0.12463209754003,
+      "learning_rate": 0.00018684210526315792,
+      "loss": 0.3191,
+      "step": 213
+    },
+    {
+      "epoch": 0.28157894736842104,
+      "grad_norm": 0.11980526105079334,
+      "learning_rate": 0.00018771929824561406,
+      "loss": 0.3238,
+      "step": 214
+    },
+    {
+      "epoch": 0.28289473684210525,
+      "grad_norm": 0.13443132852719006,
+      "learning_rate": 0.0001885964912280702,
+      "loss": 0.3409,
+      "step": 215
+    },
+    {
+      "epoch": 0.28421052631578947,
+      "grad_norm": 0.12477177793237498,
+      "learning_rate": 0.00018947368421052632,
+      "loss": 0.334,
+      "step": 216
+    },
+    {
+      "epoch": 0.2855263157894737,
+      "grad_norm": 0.12434521968333248,
+      "learning_rate": 0.00019035087719298247,
+      "loss": 0.3301,
+      "step": 217
+    },
+    {
+      "epoch": 0.2868421052631579,
+      "grad_norm": 0.12926013045173962,
+      "learning_rate": 0.0001912280701754386,
+      "loss": 0.3403,
+      "step": 218
+    },
+    {
+      "epoch": 0.2881578947368421,
+      "grad_norm": 0.12767169045744933,
+      "learning_rate": 0.00019210526315789473,
+      "loss": 0.3285,
+      "step": 219
+    },
+    {
+      "epoch": 0.2894736842105263,
+      "grad_norm": 0.12496998663175744,
+      "learning_rate": 0.00019298245614035088,
+      "loss": 0.3485,
+      "step": 220
+    },
+    {
+      "epoch": 0.29078947368421054,
+      "grad_norm": 0.12044174136139364,
+      "learning_rate": 0.00019385964912280702,
+      "loss": 0.3264,
+      "step": 221
+    },
+    {
+      "epoch": 0.29210526315789476,
+      "grad_norm": 0.13199190001366287,
+      "learning_rate": 0.00019473684210526317,
+      "loss": 0.3422,
+      "step": 222
+    },
+    {
+      "epoch": 0.29342105263157897,
+      "grad_norm": 0.1201369924836855,
+      "learning_rate": 0.0001956140350877193,
+      "loss": 0.3244,
+      "step": 223
+    },
+    {
+      "epoch": 0.29473684210526313,
+      "grad_norm": 0.11423784430905021,
+      "learning_rate": 0.00019649122807017543,
+      "loss": 0.3347,
+      "step": 224
+    },
+    {
+      "epoch": 0.29605263157894735,
+      "grad_norm": 0.12466053172472819,
+      "learning_rate": 0.00019736842105263157,
+      "loss": 0.3309,
+      "step": 225
+    },
+    {
+      "epoch": 0.29736842105263156,
+      "grad_norm": 0.12442689437527811,
+      "learning_rate": 0.00019824561403508772,
+      "loss": 0.3354,
+      "step": 226
+    },
+    {
+      "epoch": 0.2986842105263158,
+      "grad_norm": 0.11914676425566721,
+      "learning_rate": 0.00019912280701754386,
+      "loss": 0.3367,
+      "step": 227
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12114432656113984,
+      "learning_rate": 0.0002,
+      "loss": 0.3225,
+      "step": 228
+    },
+    {
+      "epoch": 0.3013157894736842,
+      "grad_norm": 0.12708865662155097,
+      "learning_rate": 0.00019999988280341633,
+      "loss": 0.3267,
+      "step": 229
+    },
+    {
+      "epoch": 0.3026315789473684,
+      "grad_norm": 0.12093461278364943,
+      "learning_rate": 0.00019999953121394002,
+      "loss": 0.3341,
+      "step": 230
+    },
+    {
+      "epoch": 0.30394736842105263,
+      "grad_norm": 0.1259425626165656,
+      "learning_rate": 0.00019999894523239515,
+      "loss": 0.3352,
+      "step": 231
+    },
+    {
+      "epoch": 0.30526315789473685,
+      "grad_norm": 0.11881152231373332,
+      "learning_rate": 0.00019999812486015523,
+      "loss": 0.3263,
+      "step": 232
+    },
+    {
+      "epoch": 0.30657894736842106,
+      "grad_norm": 0.13525301351060123,
+      "learning_rate": 0.00019999707009914317,
+      "loss": 0.3405,
+      "step": 233
+    },
+    {
+      "epoch": 0.3078947368421053,
+      "grad_norm": 0.12155988934380389,
+      "learning_rate": 0.00019999578095183124,
+      "loss": 0.3264,
+      "step": 234
+    },
+    {
+      "epoch": 0.3092105263157895,
+      "grad_norm": 0.12309971712704007,
+      "learning_rate": 0.00019999425742124114,
+      "loss": 0.3126,
+      "step": 235
+    },
+    {
+      "epoch": 0.3105263157894737,
+      "grad_norm": 0.12559904653470885,
+      "learning_rate": 0.00019999249951094388,
+      "loss": 0.3436,
+      "step": 236
+    },
+    {
+      "epoch": 0.3118421052631579,
+      "grad_norm": 0.11464091226244318,
+      "learning_rate": 0.00019999050722505993,
+      "loss": 0.3212,
+      "step": 237
+    },
+    {
+      "epoch": 0.3131578947368421,
+      "grad_norm": 0.11798628745307271,
+      "learning_rate": 0.000199988280568259,
+      "loss": 0.3187,
+      "step": 238
+    },
+    {
+      "epoch": 0.3144736842105263,
+      "grad_norm": 0.11937212399519956,
+      "learning_rate": 0.00019998581954576032,
+      "loss": 0.3261,
+      "step": 239
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.10973860755841888,
+      "learning_rate": 0.00019998312416333227,
+      "loss": 0.3097,
+      "step": 240
+    },
+    {
+      "epoch": 0.3171052631578947,
+      "grad_norm": 0.11870178537603401,
+      "learning_rate": 0.00019998019442729273,
+      "loss": 0.3264,
+      "step": 241
+    },
+    {
+      "epoch": 0.31842105263157894,
+      "grad_norm": 0.11706254926047537,
+      "learning_rate": 0.0001999770303445087,
+      "loss": 0.3276,
+      "step": 242
+    },
+    {
+      "epoch": 0.31973684210526315,
+      "grad_norm": 0.12369716377452905,
+      "learning_rate": 0.00019997363192239664,
+      "loss": 0.3178,
+      "step": 243
+    },
+    {
+      "epoch": 0.32105263157894737,
+      "grad_norm": 0.10989749847863158,
+      "learning_rate": 0.0001999699991689222,
+      "loss": 0.3126,
+      "step": 244
+    },
+    {
+      "epoch": 0.3223684210526316,
+      "grad_norm": 0.11488770235752266,
+      "learning_rate": 0.00019996613209260033,
+      "loss": 0.3305,
+      "step": 245
+    },
+    {
+      "epoch": 0.3236842105263158,
+      "grad_norm": 0.11550955957920862,
+      "learning_rate": 0.00019996203070249516,
+      "loss": 0.3395,
+      "step": 246
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 0.10896874824994508,
+      "learning_rate": 0.0001999576950082201,
+      "loss": 0.3266,
+      "step": 247
+    },
+    {
+      "epoch": 0.3263157894736842,
+      "grad_norm": 0.11429922245859891,
+      "learning_rate": 0.00019995312501993765,
+      "loss": 0.3241,
+      "step": 248
+    },
+    {
+      "epoch": 0.32763157894736844,
+      "grad_norm": 0.11185855088004702,
+      "learning_rate": 0.00019994832074835963,
+      "loss": 0.3215,
+      "step": 249
+    },
+    {
+      "epoch": 0.32894736842105265,
+      "grad_norm": 0.10618407427160971,
+      "learning_rate": 0.00019994328220474688,
+      "loss": 0.3114,
+      "step": 250
+    },
+    {
+      "epoch": 0.33026315789473687,
+      "grad_norm": 0.10827179543543274,
+      "learning_rate": 0.00019993800940090942,
+      "loss": 0.3275,
+      "step": 251
+    },
+    {
+      "epoch": 0.33157894736842103,
+      "grad_norm": 0.11600822718215488,
+      "learning_rate": 0.00019993250234920636,
+      "loss": 0.3205,
+      "step": 252
+    },
+    {
+      "epoch": 0.33289473684210524,
+      "grad_norm": 0.11660805309271545,
+      "learning_rate": 0.00019992676106254584,
+      "loss": 0.338,
+      "step": 253
+    },
+    {
+      "epoch": 0.33421052631578946,
+      "grad_norm": 0.11992791956747861,
+      "learning_rate": 0.00019992078555438502,
+      "loss": 0.33,
+      "step": 254
+    },
+    {
+      "epoch": 0.3355263157894737,
+      "grad_norm": 0.11271215008229082,
+      "learning_rate": 0.0001999145758387301,
+      "loss": 0.3312,
+      "step": 255
+    },
+    {
+      "epoch": 0.3368421052631579,
+      "grad_norm": 0.10842995849313372,
+      "learning_rate": 0.00019990813193013625,
+      "loss": 0.3257,
+      "step": 256
+    },
+    {
+      "epoch": 0.3381578947368421,
+      "grad_norm": 0.10874370438818506,
+      "learning_rate": 0.0001999014538437075,
+      "loss": 0.3128,
+      "step": 257
+    },
+    {
+      "epoch": 0.3394736842105263,
+      "grad_norm": 0.11809880440680368,
+      "learning_rate": 0.0001998945415950969,
+      "loss": 0.3211,
+      "step": 258
+    },
+    {
+      "epoch": 0.34078947368421053,
+      "grad_norm": 0.11661765420730584,
+      "learning_rate": 0.00019988739520050618,
+      "loss": 0.3148,
+      "step": 259
+    },
+    {
+      "epoch": 0.34210526315789475,
+      "grad_norm": 0.10343508953195514,
+      "learning_rate": 0.0001998800146766861,
+      "loss": 0.3217,
+      "step": 260
+    },
+    {
+      "epoch": 0.34342105263157896,
+      "grad_norm": 0.11172563699747126,
+      "learning_rate": 0.0001998724000409361,
+      "loss": 0.3098,
+      "step": 261
+    },
+    {
+      "epoch": 0.3447368421052632,
+      "grad_norm": 0.11124141647463505,
+      "learning_rate": 0.00019986455131110428,
+      "loss": 0.3177,
+      "step": 262
+    },
+    {
+      "epoch": 0.3460526315789474,
+      "grad_norm": 0.1049604206863391,
+      "learning_rate": 0.00019985646850558764,
+      "loss": 0.3202,
+      "step": 263
+    },
+    {
+      "epoch": 0.3473684210526316,
+      "grad_norm": 0.11674285217463065,
+      "learning_rate": 0.00019984815164333163,
+      "loss": 0.3325,
+      "step": 264
+    },
+    {
+      "epoch": 0.34868421052631576,
+      "grad_norm": 0.10939260536399624,
+      "learning_rate": 0.00019983960074383046,
+      "loss": 0.3135,
+      "step": 265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11228143225921142,
+      "learning_rate": 0.00019983081582712685,
+      "loss": 0.3184,
+      "step": 266
+    },
+    {
+      "epoch": 0.3513157894736842,
+      "grad_norm": 0.11736606638037488,
+      "learning_rate": 0.000199821796913812,
+      "loss": 0.3228,
+      "step": 267
+    },
+    {
+      "epoch": 0.3526315789473684,
+      "grad_norm": 0.10870195278237296,
+      "learning_rate": 0.00019981254402502566,
+      "loss": 0.3162,
+      "step": 268
+    },
+    {
+      "epoch": 0.3539473684210526,
+      "grad_norm": 0.10319062608987703,
+      "learning_rate": 0.000199803057182456,
+      "loss": 0.3088,
+      "step": 269
+    },
+    {
+      "epoch": 0.35526315789473684,
+      "grad_norm": 0.11157466968877176,
+      "learning_rate": 0.00019979333640833947,
+      "loss": 0.3157,
+      "step": 270
+    },
+    {
+      "epoch": 0.35657894736842105,
+      "grad_norm": 0.11150128423279522,
+      "learning_rate": 0.00019978338172546093,
+      "loss": 0.3101,
+      "step": 271
+    },
+    {
+      "epoch": 0.35789473684210527,
+      "grad_norm": 0.1085421577182214,
+      "learning_rate": 0.0001997731931571535,
+      "loss": 0.3126,
+      "step": 272
+    },
+    {
+      "epoch": 0.3592105263157895,
+      "grad_norm": 0.10475914894230995,
+      "learning_rate": 0.00019976277072729845,
+      "loss": 0.3224,
+      "step": 273
+    },
+    {
+      "epoch": 0.3605263157894737,
+      "grad_norm": 0.1082831141347499,
+      "learning_rate": 0.00019975211446032526,
+      "loss": 0.3082,
+      "step": 274
+    },
+    {
+      "epoch": 0.3618421052631579,
+      "grad_norm": 0.10740486742911703,
+      "learning_rate": 0.0001997412243812115,
+      "loss": 0.3257,
+      "step": 275
+    },
+    {
+      "epoch": 0.3631578947368421,
+      "grad_norm": 0.1020992105394808,
+      "learning_rate": 0.00019973010051548275,
+      "loss": 0.3156,
+      "step": 276
+    },
+    {
+      "epoch": 0.36447368421052634,
+      "grad_norm": 0.10963434549263805,
+      "learning_rate": 0.0001997187428892126,
+      "loss": 0.312,
+      "step": 277
+    },
+    {
+      "epoch": 0.36578947368421055,
+      "grad_norm": 0.10738473907360331,
+      "learning_rate": 0.00019970715152902254,
+      "loss": 0.3264,
+      "step": 278
+    },
+    {
+      "epoch": 0.3671052631578947,
+      "grad_norm": 0.10172040132714297,
+      "learning_rate": 0.00019969532646208195,
+      "loss": 0.3,
+      "step": 279
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 0.10348844172942914,
+      "learning_rate": 0.00019968326771610797,
+      "loss": 0.3116,
+      "step": 280
+    },
+    {
+      "epoch": 0.36973684210526314,
+      "grad_norm": 0.11435260003026458,
+      "learning_rate": 0.00019967097531936546,
+      "loss": 0.3196,
+      "step": 281
+    },
+    {
+      "epoch": 0.37105263157894736,
+      "grad_norm": 0.1121081703641299,
+      "learning_rate": 0.000199658449300667,
+      "loss": 0.3226,
+      "step": 282
+    },
+    {
+      "epoch": 0.37236842105263157,
+      "grad_norm": 0.1067537506801231,
+      "learning_rate": 0.00019964568968937267,
+      "loss": 0.3027,
+      "step": 283
+    },
+    {
+      "epoch": 0.3736842105263158,
+      "grad_norm": 0.11350941919576825,
+      "learning_rate": 0.00019963269651539017,
+      "loss": 0.3328,
+      "step": 284
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.10315345770730597,
+      "learning_rate": 0.00019961946980917456,
+      "loss": 0.3007,
+      "step": 285
+    },
+    {
+      "epoch": 0.3763157894736842,
+      "grad_norm": 0.11217086438246132,
+      "learning_rate": 0.0001996060096017284,
+      "loss": 0.307,
+      "step": 286
+    },
+    {
+      "epoch": 0.37763157894736843,
+      "grad_norm": 0.10830051369447247,
+      "learning_rate": 0.00019959231592460143,
+      "loss": 0.3128,
+      "step": 287
+    },
+    {
+      "epoch": 0.37894736842105264,
+      "grad_norm": 0.11069575850173559,
+      "learning_rate": 0.00019957838880989078,
+      "loss": 0.3096,
+      "step": 288
+    },
+    {
+      "epoch": 0.38026315789473686,
+      "grad_norm": 0.1085222475355183,
+      "learning_rate": 0.00019956422829024055,
+      "loss": 0.315,
+      "step": 289
+    },
+    {
+      "epoch": 0.3815789473684211,
+      "grad_norm": 0.09952894778052519,
+      "learning_rate": 0.0001995498343988421,
+      "loss": 0.3101,
+      "step": 290
+    },
+    {
+      "epoch": 0.3828947368421053,
+      "grad_norm": 0.11065531030038152,
+      "learning_rate": 0.00019953520716943371,
+      "loss": 0.3088,
+      "step": 291
+    },
+    {
+      "epoch": 0.38421052631578945,
+      "grad_norm": 0.11350469293630033,
+      "learning_rate": 0.00019952034663630062,
+      "loss": 0.318,
+      "step": 292
+    },
+    {
+      "epoch": 0.38552631578947366,
+      "grad_norm": 0.11212904813152226,
+      "learning_rate": 0.00019950525283427491,
+      "loss": 0.3172,
+      "step": 293
+    },
+    {
+      "epoch": 0.3868421052631579,
+      "grad_norm": 0.11122338935927506,
+      "learning_rate": 0.00019948992579873538,
+      "loss": 0.3058,
+      "step": 294
+    },
+    {
+      "epoch": 0.3881578947368421,
+      "grad_norm": 0.11166549208778644,
+      "learning_rate": 0.0001994743655656076,
+      "loss": 0.3014,
+      "step": 295
+    },
+    {
+      "epoch": 0.3894736842105263,
+      "grad_norm": 0.11276484187285273,
+      "learning_rate": 0.00019945857217136363,
+      "loss": 0.3177,
+      "step": 296
+    },
+    {
+      "epoch": 0.3907894736842105,
+      "grad_norm": 0.111642408366338,
+      "learning_rate": 0.00019944254565302217,
+      "loss": 0.3077,
+      "step": 297
+    },
+    {
+      "epoch": 0.39210526315789473,
+      "grad_norm": 0.12415010602542942,
+      "learning_rate": 0.00019942628604814825,
+      "loss": 0.3109,
+      "step": 298
+    },
+    {
+      "epoch": 0.39342105263157895,
+      "grad_norm": 0.10701081689961081,
+      "learning_rate": 0.00019940979339485332,
+      "loss": 0.3027,
+      "step": 299
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 0.1036095978911461,
+      "learning_rate": 0.00019939306773179497,
+      "loss": 0.3155,
+      "step": 300
+    },
+    {
+      "epoch": 0.3960526315789474,
+      "grad_norm": 0.12314325764301473,
+      "learning_rate": 0.00019937610909817702,
+      "loss": 0.3197,
+      "step": 301
+    },
+    {
+      "epoch": 0.3973684210526316,
+      "grad_norm": 0.10965174611997063,
+      "learning_rate": 0.0001993589175337494,
+      "loss": 0.3055,
+      "step": 302
+    },
+    {
+      "epoch": 0.3986842105263158,
+      "grad_norm": 0.10691327868633747,
+      "learning_rate": 0.00019934149307880791,
+      "loss": 0.3047,
+      "step": 303
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11048488675527199,
+      "learning_rate": 0.00019932383577419432,
+      "loss": 0.3162,
+      "step": 304
+    },
+    {
+      "epoch": 0.40131578947368424,
+      "grad_norm": 0.115786509511319,
+      "learning_rate": 0.0001993059456612961,
+      "loss": 0.2998,
+      "step": 305
+    },
+    {
+      "epoch": 0.4026315789473684,
+      "grad_norm": 0.10197052941221454,
+      "learning_rate": 0.0001992878227820465,
+      "loss": 0.3167,
+      "step": 306
+    },
+    {
+      "epoch": 0.4039473684210526,
+      "grad_norm": 0.10818823997320136,
+      "learning_rate": 0.00019926946717892428,
+      "loss": 0.3175,
+      "step": 307
+    },
+    {
+      "epoch": 0.4052631578947368,
+      "grad_norm": 0.11647859718654414,
+      "learning_rate": 0.00019925087889495374,
+      "loss": 0.3223,
+      "step": 308
+    },
+    {
+      "epoch": 0.40657894736842104,
+      "grad_norm": 0.10764841122975913,
+      "learning_rate": 0.0001992320579737045,
+      "loss": 0.3134,
+      "step": 309
+    },
+    {
+      "epoch": 0.40789473684210525,
+      "grad_norm": 0.10744012941703222,
+      "learning_rate": 0.0001992130044592916,
+      "loss": 0.3072,
+      "step": 310
+    },
+    {
+      "epoch": 0.40921052631578947,
+      "grad_norm": 0.11064482894831129,
+      "learning_rate": 0.00019919371839637512,
+      "loss": 0.315,
+      "step": 311
+    },
+    {
+      "epoch": 0.4105263157894737,
+      "grad_norm": 0.10891430893886998,
+      "learning_rate": 0.00019917419983016025,
+      "loss": 0.3016,
+      "step": 312
+    },
+    {
+      "epoch": 0.4118421052631579,
+      "grad_norm": 0.09966222612672655,
+      "learning_rate": 0.0001991544488063972,
+      "loss": 0.3104,
+      "step": 313
+    },
+    {
+      "epoch": 0.4131578947368421,
+      "grad_norm": 0.09948289810327957,
+      "learning_rate": 0.00019913446537138106,
+      "loss": 0.3199,
+      "step": 314
+    },
+    {
+      "epoch": 0.4144736842105263,
+      "grad_norm": 0.10221948739250226,
+      "learning_rate": 0.00019911424957195158,
+      "loss": 0.3027,
+      "step": 315
+    },
+    {
+      "epoch": 0.41578947368421054,
+      "grad_norm": 0.10319916825175317,
+      "learning_rate": 0.00019909380145549324,
+      "loss": 0.2962,
+      "step": 316
+    },
+    {
+      "epoch": 0.41710526315789476,
+      "grad_norm": 0.10112207458406539,
+      "learning_rate": 0.00019907312106993503,
+      "loss": 0.3072,
+      "step": 317
+    },
+    {
+      "epoch": 0.41842105263157897,
+      "grad_norm": 0.1056216300520592,
+      "learning_rate": 0.00019905220846375032,
+      "loss": 0.3108,
+      "step": 318
+    },
+    {
+      "epoch": 0.41973684210526313,
+      "grad_norm": 0.10432503631292016,
+      "learning_rate": 0.0001990310636859569,
+      "loss": 0.3084,
+      "step": 319
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.10088521223549587,
+      "learning_rate": 0.00019900968678611666,
+      "loss": 0.3094,
+      "step": 320
+    },
+    {
+      "epoch": 0.42236842105263156,
+      "grad_norm": 0.09994245819164545,
+      "learning_rate": 0.00019898807781433555,
+      "loss": 0.3142,
+      "step": 321
+    },
+    {
+      "epoch": 0.4236842105263158,
+      "grad_norm": 0.09903460972489679,
+      "learning_rate": 0.00019896623682126355,
+      "loss": 0.307,
+      "step": 322
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 0.09914885914415125,
+      "learning_rate": 0.00019894416385809444,
+      "loss": 0.3014,
+      "step": 323
+    },
+    {
+      "epoch": 0.4263157894736842,
+      "grad_norm": 0.10122296873459234,
+      "learning_rate": 0.00019892185897656578,
+      "loss": 0.303,
+      "step": 324
+    },
+    {
+      "epoch": 0.4276315789473684,
+      "grad_norm": 0.09910489155114076,
+      "learning_rate": 0.0001988993222289587,
+      "loss": 0.3117,
+      "step": 325
+    },
+    {
+      "epoch": 0.42894736842105263,
+      "grad_norm": 0.09241116281014006,
+      "learning_rate": 0.0001988765536680977,
+      "loss": 0.2934,
+      "step": 326
+    },
+    {
+      "epoch": 0.43026315789473685,
+      "grad_norm": 0.09770357586471753,
+      "learning_rate": 0.00019885355334735082,
+      "loss": 0.317,
+      "step": 327
+    },
+    {
+      "epoch": 0.43157894736842106,
+      "grad_norm": 0.09697496010174222,
+      "learning_rate": 0.00019883032132062925,
+      "loss": 0.2998,
+      "step": 328
+    },
+    {
+      "epoch": 0.4328947368421053,
+      "grad_norm": 0.10013197876212468,
+      "learning_rate": 0.0001988068576423872,
+      "loss": 0.3105,
+      "step": 329
+    },
+    {
+      "epoch": 0.4342105263157895,
+      "grad_norm": 0.09453550444418382,
+      "learning_rate": 0.00019878316236762196,
+      "loss": 0.3091,
+      "step": 330
+    },
+    {
+      "epoch": 0.4355263157894737,
+      "grad_norm": 0.10361296309003387,
+      "learning_rate": 0.00019875923555187365,
+      "loss": 0.3117,
+      "step": 331
+    },
+    {
+      "epoch": 0.4368421052631579,
+      "grad_norm": 0.09867438093722174,
+      "learning_rate": 0.00019873507725122504,
+      "loss": 0.3105,
+      "step": 332
+    },
+    {
+      "epoch": 0.4381578947368421,
+      "grad_norm": 0.09914171961719935,
+      "learning_rate": 0.00019871068752230162,
+      "loss": 0.3299,
+      "step": 333
+    },
+    {
+      "epoch": 0.4394736842105263,
+      "grad_norm": 0.09508589565593037,
+      "learning_rate": 0.00019868606642227122,
+      "loss": 0.3057,
+      "step": 334
+    },
+    {
+      "epoch": 0.4407894736842105,
+      "grad_norm": 0.10553369850294496,
+      "learning_rate": 0.00019866121400884397,
+      "loss": 0.2968,
+      "step": 335
+    },
+    {
+      "epoch": 0.4421052631578947,
+      "grad_norm": 0.09693152875657689,
+      "learning_rate": 0.00019863613034027224,
+      "loss": 0.3157,
+      "step": 336
+    },
+    {
+      "epoch": 0.44342105263157894,
+      "grad_norm": 0.09914036757559133,
+      "learning_rate": 0.0001986108154753505,
+      "loss": 0.3075,
+      "step": 337
+    },
+    {
+      "epoch": 0.44473684210526315,
+      "grad_norm": 0.10323372097858888,
+      "learning_rate": 0.00019858526947341497,
+      "loss": 0.316,
+      "step": 338
+    },
+    {
+      "epoch": 0.44605263157894737,
+      "grad_norm": 0.1047670642802198,
+      "learning_rate": 0.0001985594923943438,
+      "loss": 0.3016,
+      "step": 339
+    },
+    {
+      "epoch": 0.4473684210526316,
+      "grad_norm": 0.10038200225737869,
+      "learning_rate": 0.00019853348429855672,
+      "loss": 0.313,
+      "step": 340
+    },
+    {
+      "epoch": 0.4486842105263158,
+      "grad_norm": 0.09995800002881376,
+      "learning_rate": 0.00019850724524701486,
+      "loss": 0.3201,
+      "step": 341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09706393223860003,
+      "learning_rate": 0.00019848077530122083,
+      "loss": 0.302,
+      "step": 342
+    },
+    {
+      "epoch": 0.4513157894736842,
+      "grad_norm": 0.10217585842187529,
+      "learning_rate": 0.0001984540745232183,
+      "loss": 0.3052,
+      "step": 343
+    },
+    {
+      "epoch": 0.45263157894736844,
+      "grad_norm": 0.09716163731483823,
+      "learning_rate": 0.00019842714297559213,
+      "loss": 0.3182,
+      "step": 344
+    },
+    {
+      "epoch": 0.45394736842105265,
+      "grad_norm": 0.09580525044842596,
+      "learning_rate": 0.000198399980721468,
+      "loss": 0.3016,
+      "step": 345
+    },
+    {
+      "epoch": 0.45526315789473687,
+      "grad_norm": 0.10216214300781075,
+      "learning_rate": 0.0001983725878245124,
+      "loss": 0.3093,
+      "step": 346
+    },
+    {
+      "epoch": 0.45657894736842103,
+      "grad_norm": 0.09748580988527811,
+      "learning_rate": 0.0001983449643489324,
+      "loss": 0.3077,
+      "step": 347
+    },
+    {
+      "epoch": 0.45789473684210524,
+      "grad_norm": 0.10212802995697862,
+      "learning_rate": 0.0001983171103594755,
+      "loss": 0.3039,
+      "step": 348
+    },
+    {
+      "epoch": 0.45921052631578946,
+      "grad_norm": 0.10018192206504958,
+      "learning_rate": 0.00019828902592142962,
+      "loss": 0.3023,
+      "step": 349
+    },
+    {
+      "epoch": 0.4605263157894737,
+      "grad_norm": 0.09425076167869935,
+      "learning_rate": 0.0001982607111006227,
+      "loss": 0.3119,
+      "step": 350
+    },
+    {
+      "epoch": 0.4618421052631579,
+      "grad_norm": 0.0967215301525356,
+      "learning_rate": 0.0001982321659634228,
+      "loss": 0.3097,
+      "step": 351
+    },
+    {
+      "epoch": 0.4631578947368421,
+      "grad_norm": 0.09622422612500942,
+      "learning_rate": 0.0001982033905767377,
+      "loss": 0.3114,
+      "step": 352
+    },
+    {
+      "epoch": 0.4644736842105263,
+      "grad_norm": 0.09360899731223273,
+      "learning_rate": 0.00019817438500801502,
+      "loss": 0.3099,
+      "step": 353
+    },
+    {
+      "epoch": 0.46578947368421053,
+      "grad_norm": 0.09987640487171251,
+      "learning_rate": 0.0001981451493252418,
+      "loss": 0.2994,
+      "step": 354
+    },
+    {
+      "epoch": 0.46710526315789475,
+      "grad_norm": 0.10065235571453063,
+      "learning_rate": 0.00019811568359694447,
+      "loss": 0.3048,
+      "step": 355
+    },
+    {
+      "epoch": 0.46842105263157896,
+      "grad_norm": 0.10208391531367922,
+      "learning_rate": 0.00019808598789218865,
+      "loss": 0.2939,
+      "step": 356
+    },
+    {
+      "epoch": 0.4697368421052632,
+      "grad_norm": 0.10402670400636525,
+      "learning_rate": 0.00019805606228057916,
+      "loss": 0.3024,
+      "step": 357
+    },
+    {
+      "epoch": 0.4710526315789474,
+      "grad_norm": 0.09930311628562899,
+      "learning_rate": 0.00019802590683225946,
+      "loss": 0.3024,
+      "step": 358
+    },
+    {
+      "epoch": 0.4723684210526316,
+      "grad_norm": 0.10262536271155684,
+      "learning_rate": 0.0001979955216179119,
+      "loss": 0.3158,
+      "step": 359
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.0943881489890398,
+      "learning_rate": 0.0001979649067087574,
+      "loss": 0.3095,
+      "step": 360
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 0.09827781548848098,
+      "learning_rate": 0.00019793406217655517,
+      "loss": 0.3084,
+      "step": 361
+    },
+    {
+      "epoch": 0.4763157894736842,
+      "grad_norm": 0.09396956618890727,
+      "learning_rate": 0.00019790298809360267,
+      "loss": 0.3108,
+      "step": 362
+    },
+    {
+      "epoch": 0.4776315789473684,
+      "grad_norm": 0.0971493552146761,
+      "learning_rate": 0.00019787168453273544,
+      "loss": 0.3234,
+      "step": 363
+    },
+    {
+      "epoch": 0.4789473684210526,
+      "grad_norm": 0.09635250868834104,
+      "learning_rate": 0.00019784015156732693,
+      "loss": 0.2961,
+      "step": 364
+    },
+    {
+      "epoch": 0.48026315789473684,
+      "grad_norm": 0.0957638945230677,
+      "learning_rate": 0.00019780838927128822,
+      "loss": 0.2999,
+      "step": 365
+    },
+    {
+      "epoch": 0.48157894736842105,
+      "grad_norm": 0.09673326599521005,
+      "learning_rate": 0.00019777639771906795,
+      "loss": 0.3065,
+      "step": 366
+    },
+    {
+      "epoch": 0.48289473684210527,
+      "grad_norm": 0.10037616383525308,
+      "learning_rate": 0.00019774417698565215,
+      "loss": 0.3155,
+      "step": 367
+    },
+    {
+      "epoch": 0.4842105263157895,
+      "grad_norm": 0.09978049327969367,
+      "learning_rate": 0.000197711727146564,
+      "loss": 0.3165,
+      "step": 368
+    },
+    {
+      "epoch": 0.4855263157894737,
+      "grad_norm": 0.09791958592311975,
+      "learning_rate": 0.00019767904827786375,
+      "loss": 0.3089,
+      "step": 369
+    },
+    {
+      "epoch": 0.4868421052631579,
+      "grad_norm": 0.09995882365072112,
+      "learning_rate": 0.00019764614045614836,
+      "loss": 0.2988,
+      "step": 370
+    },
+    {
+      "epoch": 0.4881578947368421,
+      "grad_norm": 0.09232234693055072,
+      "learning_rate": 0.0001976130037585516,
+      "loss": 0.3042,
+      "step": 371
+    },
+    {
+      "epoch": 0.48947368421052634,
+      "grad_norm": 0.09643481824906593,
+      "learning_rate": 0.00019757963826274357,
+      "loss": 0.294,
+      "step": 372
+    },
+    {
+      "epoch": 0.49078947368421055,
+      "grad_norm": 0.09697913699025175,
+      "learning_rate": 0.00019754604404693073,
+      "loss": 0.3053,
+      "step": 373
+    },
+    {
+      "epoch": 0.4921052631578947,
+      "grad_norm": 0.09907221334912318,
+      "learning_rate": 0.00019751222118985563,
+      "loss": 0.2968,
+      "step": 374
+    },
+    {
+      "epoch": 0.4934210526315789,
+      "grad_norm": 0.09502914913791907,
+      "learning_rate": 0.00019747816977079671,
+      "loss": 0.3078,
+      "step": 375
+    },
+    {
+      "epoch": 0.49473684210526314,
+      "grad_norm": 0.09936470649154033,
+      "learning_rate": 0.00019744388986956822,
+      "loss": 0.3097,
+      "step": 376
+    },
+    {
+      "epoch": 0.49605263157894736,
+      "grad_norm": 0.09802911974074544,
+      "learning_rate": 0.0001974093815665199,
+      "loss": 0.2958,
+      "step": 377
+    },
+    {
+      "epoch": 0.49736842105263157,
+      "grad_norm": 0.09898755664324496,
+      "learning_rate": 0.0001973746449425368,
+      "loss": 0.3083,
+      "step": 378
+    },
+    {
+      "epoch": 0.4986842105263158,
+      "grad_norm": 0.0961406374333812,
+      "learning_rate": 0.00019733968007903922,
+      "loss": 0.2976,
+      "step": 379
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10159753311874659,
+      "learning_rate": 0.00019730448705798239,
+      "loss": 0.3033,
+      "step": 380
+    },
+    {
+      "epoch": 0.5013157894736842,
+      "grad_norm": 0.0947441745895011,
+      "learning_rate": 0.0001972690659618564,
+      "loss": 0.2852,
+      "step": 381
+    },
+    {
+      "epoch": 0.5026315789473684,
+      "grad_norm": 0.09775748236249004,
+      "learning_rate": 0.00019723341687368583,
+      "loss": 0.2929,
+      "step": 382
+    },
+    {
+      "epoch": 0.5039473684210526,
+      "grad_norm": 0.09935691145327614,
+      "learning_rate": 0.00019719753987702978,
+      "loss": 0.3111,
+      "step": 383
+    },
+    {
+      "epoch": 0.5052631578947369,
+      "grad_norm": 0.09893488807438079,
+      "learning_rate": 0.0001971614350559814,
+      "loss": 0.3158,
+      "step": 384
+    },
+    {
+      "epoch": 0.506578947368421,
+      "grad_norm": 0.0947497912778192,
+      "learning_rate": 0.00019712510249516793,
+      "loss": 0.2921,
+      "step": 385
+    },
+    {
+      "epoch": 0.5078947368421053,
+      "grad_norm": 0.08737508895101483,
+      "learning_rate": 0.00019708854227975048,
+      "loss": 0.2927,
+      "step": 386
+    },
+    {
+      "epoch": 0.5092105263157894,
+      "grad_norm": 0.09835779795374115,
+      "learning_rate": 0.00019705175449542358,
+      "loss": 0.2976,
+      "step": 387
+    },
+    {
+      "epoch": 0.5105263157894737,
+      "grad_norm": 0.09107621488427446,
+      "learning_rate": 0.0001970147392284154,
+      "loss": 0.2814,
+      "step": 388
+    },
+    {
+      "epoch": 0.5118421052631579,
+      "grad_norm": 0.09743589157185283,
+      "learning_rate": 0.00019697749656548714,
+      "loss": 0.3003,
+      "step": 389
+    },
+    {
+      "epoch": 0.5131578947368421,
+      "grad_norm": 0.09595683633527828,
+      "learning_rate": 0.00019694002659393305,
+      "loss": 0.3033,
+      "step": 390
+    },
+    {
+      "epoch": 0.5144736842105263,
+      "grad_norm": 0.09879879361599342,
+      "learning_rate": 0.0001969023294015802,
+      "loss": 0.3133,
+      "step": 391
+    },
+    {
+      "epoch": 0.5157894736842106,
+      "grad_norm": 0.10105822465511066,
+      "learning_rate": 0.00019686440507678824,
+      "loss": 0.3103,
+      "step": 392
+    },
+    {
+      "epoch": 0.5171052631578947,
+      "grad_norm": 0.10182072904717256,
+      "learning_rate": 0.00019682625370844918,
+      "loss": 0.2978,
+      "step": 393
+    },
+    {
+      "epoch": 0.5184210526315789,
+      "grad_norm": 0.09540364410216295,
+      "learning_rate": 0.00019678787538598725,
+      "loss": 0.3243,
+      "step": 394
+    },
+    {
+      "epoch": 0.5197368421052632,
+      "grad_norm": 0.09754493311070281,
+      "learning_rate": 0.00019674927019935857,
+      "loss": 0.2957,
+      "step": 395
+    },
+    {
+      "epoch": 0.5210526315789473,
+      "grad_norm": 0.09261668278184784,
+      "learning_rate": 0.0001967104382390511,
+      "loss": 0.2909,
+      "step": 396
+    },
+    {
+      "epoch": 0.5223684210526316,
+      "grad_norm": 0.0978202914794332,
+      "learning_rate": 0.00019667137959608426,
+      "loss": 0.2875,
+      "step": 397
+    },
+    {
+      "epoch": 0.5236842105263158,
+      "grad_norm": 0.09087851403559502,
+      "learning_rate": 0.00019663209436200887,
+      "loss": 0.2959,
+      "step": 398
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 0.09362515028658314,
+      "learning_rate": 0.00019659258262890683,
+      "loss": 0.3042,
+      "step": 399
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.09012777438494002,
+      "learning_rate": 0.00019655284448939094,
+      "loss": 0.2959,
+      "step": 400
+    },
+    {
+      "epoch": 0.5276315789473685,
+      "grad_norm": 0.0960827279363291,
+      "learning_rate": 0.0001965128800366047,
+      "loss": 0.2929,
+      "step": 401
+    },
+    {
+      "epoch": 0.5289473684210526,
+      "grad_norm": 0.09300199527426295,
+      "learning_rate": 0.00019647268936422206,
+      "loss": 0.283,
+      "step": 402
+    },
+    {
+      "epoch": 0.5302631578947369,
+      "grad_norm": 0.09683305560772038,
+      "learning_rate": 0.00019643227256644716,
+      "loss": 0.3105,
+      "step": 403
+    },
+    {
+      "epoch": 0.531578947368421,
+      "grad_norm": 0.0954490202804032,
+      "learning_rate": 0.00019639162973801426,
+      "loss": 0.3023,
+      "step": 404
+    },
+    {
+      "epoch": 0.5328947368421053,
+      "grad_norm": 0.09612701147402002,
+      "learning_rate": 0.00019635076097418734,
+      "loss": 0.3176,
+      "step": 405
+    },
+    {
+      "epoch": 0.5342105263157895,
+      "grad_norm": 0.09104963315940301,
+      "learning_rate": 0.00019630966637076004,
+      "loss": 0.2997,
+      "step": 406
+    },
+    {
+      "epoch": 0.5355263157894737,
+      "grad_norm": 0.0986012989783705,
+      "learning_rate": 0.00019626834602405523,
+      "loss": 0.2964,
+      "step": 407
+    },
+    {
+      "epoch": 0.5368421052631579,
+      "grad_norm": 0.10336828609236494,
+      "learning_rate": 0.00019622680003092503,
+      "loss": 0.2886,
+      "step": 408
+    },
+    {
+      "epoch": 0.5381578947368421,
+      "grad_norm": 0.09141302481564861,
+      "learning_rate": 0.00019618502848875045,
+      "loss": 0.2898,
+      "step": 409
+    },
+    {
+      "epoch": 0.5394736842105263,
+      "grad_norm": 0.09811357001705759,
+      "learning_rate": 0.00019614303149544102,
+      "loss": 0.312,
+      "step": 410
+    },
+    {
+      "epoch": 0.5407894736842105,
+      "grad_norm": 0.09760884292596834,
+      "learning_rate": 0.00019610080914943492,
+      "loss": 0.31,
+      "step": 411
+    },
+    {
+      "epoch": 0.5421052631578948,
+      "grad_norm": 0.09299566319312146,
+      "learning_rate": 0.0001960583615496984,
+      "loss": 0.2892,
+      "step": 412
+    },
+    {
+      "epoch": 0.5434210526315789,
+      "grad_norm": 0.09898383275023284,
+      "learning_rate": 0.0001960156887957257,
+      "loss": 0.3005,
+      "step": 413
+    },
+    {
+      "epoch": 0.5447368421052632,
+      "grad_norm": 0.10036363231456787,
+      "learning_rate": 0.0001959727909875389,
+      "loss": 0.2882,
+      "step": 414
+    },
+    {
+      "epoch": 0.5460526315789473,
+      "grad_norm": 0.08979092246398529,
+      "learning_rate": 0.00019592966822568753,
+      "loss": 0.301,
+      "step": 415
+    },
+    {
+      "epoch": 0.5473684210526316,
+      "grad_norm": 0.09422502905792629,
+      "learning_rate": 0.00019588632061124837,
+      "loss": 0.2851,
+      "step": 416
+    },
+    {
+      "epoch": 0.5486842105263158,
+      "grad_norm": 0.08769883690329003,
+      "learning_rate": 0.0001958427482458253,
+      "loss": 0.2874,
+      "step": 417
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10126724866744835,
+      "learning_rate": 0.0001957989512315489,
+      "loss": 0.3068,
+      "step": 418
+    },
+    {
+      "epoch": 0.5513157894736842,
+      "grad_norm": 0.09755145401929033,
+      "learning_rate": 0.00019575492967107642,
+      "loss": 0.3072,
+      "step": 419
+    },
+    {
+      "epoch": 0.5526315789473685,
+      "grad_norm": 0.09089099824025165,
+      "learning_rate": 0.00019571068366759143,
+      "loss": 0.2891,
+      "step": 420
+    },
+    {
+      "epoch": 0.5539473684210526,
+      "grad_norm": 0.0961002056223822,
+      "learning_rate": 0.00019566621332480348,
+      "loss": 0.2935,
+      "step": 421
+    },
+    {
+      "epoch": 0.5552631578947368,
+      "grad_norm": 0.09426174969645366,
+      "learning_rate": 0.00019562151874694803,
+      "loss": 0.2925,
+      "step": 422
+    },
+    {
+      "epoch": 0.5565789473684211,
+      "grad_norm": 0.09928190504185619,
+      "learning_rate": 0.00019557660003878614,
+      "loss": 0.3055,
+      "step": 423
+    },
+    {
+      "epoch": 0.5578947368421052,
+      "grad_norm": 0.09531727011576784,
+      "learning_rate": 0.00019553145730560415,
+      "loss": 0.2916,
+      "step": 424
+    },
+    {
+      "epoch": 0.5592105263157895,
+      "grad_norm": 0.09655059432651882,
+      "learning_rate": 0.00019548609065321356,
+      "loss": 0.3058,
+      "step": 425
+    },
+    {
+      "epoch": 0.5605263157894737,
+      "grad_norm": 0.09707096390323401,
+      "learning_rate": 0.00019544050018795075,
+      "loss": 0.2873,
+      "step": 426
+    },
+    {
+      "epoch": 0.5618421052631579,
+      "grad_norm": 0.09428953984002672,
+      "learning_rate": 0.00019539468601667662,
+      "loss": 0.298,
+      "step": 427
+    },
+    {
+      "epoch": 0.5631578947368421,
+      "grad_norm": 0.09658263944025584,
+      "learning_rate": 0.0001953486482467764,
+      "loss": 0.3004,
+      "step": 428
+    },
+    {
+      "epoch": 0.5644736842105263,
+      "grad_norm": 0.10052098718968268,
+      "learning_rate": 0.00019530238698615957,
+      "loss": 0.2975,
+      "step": 429
+    },
+    {
+      "epoch": 0.5657894736842105,
+      "grad_norm": 0.09350068437861885,
+      "learning_rate": 0.00019525590234325933,
+      "loss": 0.2977,
+      "step": 430
+    },
+    {
+      "epoch": 0.5671052631578948,
+      "grad_norm": 0.09265833701591007,
+      "learning_rate": 0.00019520919442703245,
+      "loss": 0.2941,
+      "step": 431
+    },
+    {
+      "epoch": 0.5684210526315789,
+      "grad_norm": 0.0914948295571144,
+      "learning_rate": 0.0001951622633469592,
+      "loss": 0.3032,
+      "step": 432
+    },
+    {
+      "epoch": 0.5697368421052632,
+      "grad_norm": 0.09538108350867988,
+      "learning_rate": 0.00019511510921304273,
+      "loss": 0.3057,
+      "step": 433
+    },
+    {
+      "epoch": 0.5710526315789474,
+      "grad_norm": 0.09388517597881102,
+      "learning_rate": 0.00019506773213580917,
+      "loss": 0.2974,
+      "step": 434
+    },
+    {
+      "epoch": 0.5723684210526315,
+      "grad_norm": 0.09853399504304618,
+      "learning_rate": 0.00019502013222630712,
+      "loss": 0.2831,
+      "step": 435
+    },
+    {
+      "epoch": 0.5736842105263158,
+      "grad_norm": 0.09228736628869874,
+      "learning_rate": 0.00019497230959610756,
+      "loss": 0.2885,
+      "step": 436
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 0.09786678573184258,
+      "learning_rate": 0.0001949242643573034,
+      "loss": 0.3082,
+      "step": 437
+    },
+    {
+      "epoch": 0.5763157894736842,
+      "grad_norm": 0.09190204992934663,
+      "learning_rate": 0.00019487599662250943,
+      "loss": 0.2974,
+      "step": 438
+    },
+    {
+      "epoch": 0.5776315789473684,
+      "grad_norm": 0.09838875478146845,
+      "learning_rate": 0.00019482750650486193,
+      "loss": 0.2869,
+      "step": 439
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.09080464891287887,
+      "learning_rate": 0.00019477879411801844,
+      "loss": 0.2999,
+      "step": 440
+    },
+    {
+      "epoch": 0.5802631578947368,
+      "grad_norm": 0.094818410464559,
+      "learning_rate": 0.0001947298595761574,
+      "loss": 0.2948,
+      "step": 441
+    },
+    {
+      "epoch": 0.5815789473684211,
+      "grad_norm": 0.09229949003285352,
+      "learning_rate": 0.00019468070299397808,
+      "loss": 0.2851,
+      "step": 442
+    },
+    {
+      "epoch": 0.5828947368421052,
+      "grad_norm": 0.0946200275878873,
+      "learning_rate": 0.0001946313244867002,
+      "loss": 0.3035,
+      "step": 443
+    },
+    {
+      "epoch": 0.5842105263157895,
+      "grad_norm": 0.0972452681474162,
+      "learning_rate": 0.00019458172417006347,
+      "loss": 0.2921,
+      "step": 444
+    },
+    {
+      "epoch": 0.5855263157894737,
+      "grad_norm": 0.09168097969668783,
+      "learning_rate": 0.00019453190216032776,
+      "loss": 0.2965,
+      "step": 445
+    },
+    {
+      "epoch": 0.5868421052631579,
+      "grad_norm": 0.09238240620410484,
+      "learning_rate": 0.00019448185857427242,
+      "loss": 0.2965,
+      "step": 446
+    },
+    {
+      "epoch": 0.5881578947368421,
+      "grad_norm": 0.10748289837029415,
+      "learning_rate": 0.00019443159352919623,
+      "loss": 0.2912,
+      "step": 447
+    },
+    {
+      "epoch": 0.5894736842105263,
+      "grad_norm": 0.09871636341616816,
+      "learning_rate": 0.00019438110714291694,
+      "loss": 0.3111,
+      "step": 448
+    },
+    {
+      "epoch": 0.5907894736842105,
+      "grad_norm": 0.08861315777205786,
+      "learning_rate": 0.00019433039953377127,
+      "loss": 0.2891,
+      "step": 449
+    },
+    {
+      "epoch": 0.5921052631578947,
+      "grad_norm": 0.09570467370902319,
+      "learning_rate": 0.00019427947082061432,
+      "loss": 0.3046,
+      "step": 450
+    },
+    {
+      "epoch": 0.593421052631579,
+      "grad_norm": 0.0882411637366609,
+      "learning_rate": 0.00019422832112281962,
+      "loss": 0.2948,
+      "step": 451
+    },
+    {
+      "epoch": 0.5947368421052631,
+      "grad_norm": 0.08681692074605786,
+      "learning_rate": 0.00019417695056027844,
+      "loss": 0.2975,
+      "step": 452
+    },
+    {
+      "epoch": 0.5960526315789474,
+      "grad_norm": 0.0894965926644007,
+      "learning_rate": 0.00019412535925339997,
+      "loss": 0.2982,
+      "step": 453
+    },
+    {
+      "epoch": 0.5973684210526315,
+      "grad_norm": 0.08949478378197007,
+      "learning_rate": 0.00019407354732311064,
+      "loss": 0.2923,
+      "step": 454
+    },
+    {
+      "epoch": 0.5986842105263158,
+      "grad_norm": 0.0952879006180906,
+      "learning_rate": 0.0001940215148908541,
+      "loss": 0.2959,
+      "step": 455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09341215938159328,
+      "learning_rate": 0.00019396926207859084,
+      "loss": 0.2987,
+      "step": 456
+    },
+    {
+      "epoch": 0.6013157894736842,
+      "grad_norm": 0.08989810174624617,
+      "learning_rate": 0.00019391678900879786,
+      "loss": 0.2991,
+      "step": 457
+    },
+    {
+      "epoch": 0.6026315789473684,
+      "grad_norm": 0.09052527717801154,
+      "learning_rate": 0.00019386409580446844,
+      "loss": 0.2841,
+      "step": 458
+    },
+    {
+      "epoch": 0.6039473684210527,
+      "grad_norm": 0.09246686105449765,
+      "learning_rate": 0.00019381118258911186,
+      "loss": 0.2998,
+      "step": 459
+    },
+    {
+      "epoch": 0.6052631578947368,
+      "grad_norm": 0.09632503586933118,
+      "learning_rate": 0.00019375804948675306,
+      "loss": 0.3075,
+      "step": 460
+    },
+    {
+      "epoch": 0.6065789473684211,
+      "grad_norm": 0.09476157150253976,
+      "learning_rate": 0.00019370469662193248,
+      "loss": 0.2981,
+      "step": 461
+    },
+    {
+      "epoch": 0.6078947368421053,
+      "grad_norm": 0.08742707617370972,
+      "learning_rate": 0.0001936511241197055,
+      "loss": 0.2805,
+      "step": 462
+    },
+    {
+      "epoch": 0.6092105263157894,
+      "grad_norm": 0.09283003181432853,
+      "learning_rate": 0.00019359733210564244,
+      "loss": 0.2955,
+      "step": 463
+    },
+    {
+      "epoch": 0.6105263157894737,
+      "grad_norm": 0.09458543714491262,
+      "learning_rate": 0.0001935433207058281,
+      "loss": 0.2949,
+      "step": 464
+    },
+    {
+      "epoch": 0.6118421052631579,
+      "grad_norm": 0.08784515458042345,
+      "learning_rate": 0.00019348909004686152,
+      "loss": 0.3033,
+      "step": 465
+    },
+    {
+      "epoch": 0.6131578947368421,
+      "grad_norm": 0.08813027640751597,
+      "learning_rate": 0.00019343464025585563,
+      "loss": 0.2949,
+      "step": 466
+    },
+    {
+      "epoch": 0.6144736842105263,
+      "grad_norm": 0.09671795564332149,
+      "learning_rate": 0.00019337997146043708,
+      "loss": 0.3022,
+      "step": 467
+    },
+    {
+      "epoch": 0.6157894736842106,
+      "grad_norm": 0.09555526827290024,
+      "learning_rate": 0.0001933250837887457,
+      "loss": 0.2956,
+      "step": 468
+    },
+    {
+      "epoch": 0.6171052631578947,
+      "grad_norm": 0.08589539494059092,
+      "learning_rate": 0.00019326997736943455,
+      "loss": 0.3052,
+      "step": 469
+    },
+    {
+      "epoch": 0.618421052631579,
+      "grad_norm": 0.08999975458012179,
+      "learning_rate": 0.00019321465233166924,
+      "loss": 0.2949,
+      "step": 470
+    },
+    {
+      "epoch": 0.6197368421052631,
+      "grad_norm": 0.09506140617593234,
+      "learning_rate": 0.0001931591088051279,
+      "loss": 0.3151,
+      "step": 471
+    },
+    {
+      "epoch": 0.6210526315789474,
+      "grad_norm": 0.09611958378062155,
+      "learning_rate": 0.00019310334692000075,
+      "loss": 0.2907,
+      "step": 472
+    },
+    {
+      "epoch": 0.6223684210526316,
+      "grad_norm": 0.09078308388388881,
+      "learning_rate": 0.00019304736680698988,
+      "loss": 0.2969,
+      "step": 473
+    },
+    {
+      "epoch": 0.6236842105263158,
+      "grad_norm": 0.09080240341624422,
+      "learning_rate": 0.0001929911685973088,
+      "loss": 0.2969,
+      "step": 474
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.08939149414674773,
+      "learning_rate": 0.00019293475242268223,
+      "loss": 0.2849,
+      "step": 475
+    },
+    {
+      "epoch": 0.6263157894736842,
+      "grad_norm": 0.08769153124531658,
+      "learning_rate": 0.00019287811841534595,
+      "loss": 0.2857,
+      "step": 476
+    },
+    {
+      "epoch": 0.6276315789473684,
+      "grad_norm": 0.09345365417023815,
+      "learning_rate": 0.00019282126670804614,
+      "loss": 0.3044,
+      "step": 477
+    },
+    {
+      "epoch": 0.6289473684210526,
+      "grad_norm": 0.09137801449382149,
+      "learning_rate": 0.00019276419743403933,
+      "loss": 0.2993,
+      "step": 478
+    },
+    {
+      "epoch": 0.6302631578947369,
+      "grad_norm": 0.090592310213129,
+      "learning_rate": 0.00019270691072709195,
+      "loss": 0.3005,
+      "step": 479
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.09174534803218672,
+      "learning_rate": 0.00019264940672148018,
+      "loss": 0.3059,
+      "step": 480
+    },
+    {
+      "epoch": 0.6328947368421053,
+      "grad_norm": 0.0871835682901987,
+      "learning_rate": 0.00019259168555198948,
+      "loss": 0.286,
+      "step": 481
+    },
+    {
+      "epoch": 0.6342105263157894,
+      "grad_norm": 0.0892410180390742,
+      "learning_rate": 0.0001925337473539143,
+      "loss": 0.2788,
+      "step": 482
+    },
+    {
+      "epoch": 0.6355263157894737,
+      "grad_norm": 0.09248088501840468,
+      "learning_rate": 0.00019247559226305785,
+      "loss": 0.2948,
+      "step": 483
+    },
+    {
+      "epoch": 0.6368421052631579,
+      "grad_norm": 0.09098010321941165,
+      "learning_rate": 0.00019241722041573166,
+      "loss": 0.3028,
+      "step": 484
+    },
+    {
+      "epoch": 0.6381578947368421,
+      "grad_norm": 0.0896570234030572,
+      "learning_rate": 0.00019235863194875532,
+      "loss": 0.2832,
+      "step": 485
+    },
+    {
+      "epoch": 0.6394736842105263,
+      "grad_norm": 0.08779926165652006,
+      "learning_rate": 0.0001922998269994563,
+      "loss": 0.2987,
+      "step": 486
+    },
+    {
+      "epoch": 0.6407894736842106,
+      "grad_norm": 0.09365677051236368,
+      "learning_rate": 0.00019224080570566927,
+      "loss": 0.2912,
+      "step": 487
+    },
+    {
+      "epoch": 0.6421052631578947,
+      "grad_norm": 0.08927856303791634,
+      "learning_rate": 0.0001921815682057362,
+      "loss": 0.2966,
+      "step": 488
+    },
+    {
+      "epoch": 0.6434210526315789,
+      "grad_norm": 0.0879241771326983,
+      "learning_rate": 0.00019212211463850567,
+      "loss": 0.284,
+      "step": 489
+    },
+    {
+      "epoch": 0.6447368421052632,
+      "grad_norm": 0.0878734974007894,
+      "learning_rate": 0.00019206244514333282,
+      "loss": 0.2834,
+      "step": 490
+    },
+    {
+      "epoch": 0.6460526315789473,
+      "grad_norm": 0.08899613109416632,
+      "learning_rate": 0.00019200255986007885,
+      "loss": 0.2905,
+      "step": 491
+    },
+    {
+      "epoch": 0.6473684210526316,
+      "grad_norm": 0.0868302957136876,
+      "learning_rate": 0.0001919424589291108,
+      "loss": 0.2994,
+      "step": 492
+    },
+    {
+      "epoch": 0.6486842105263158,
+      "grad_norm": 0.08873110421371189,
+      "learning_rate": 0.0001918821424913011,
+      "loss": 0.3054,
+      "step": 493
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08729340226342107,
+      "learning_rate": 0.00019182161068802741,
+      "loss": 0.2926,
+      "step": 494
+    },
+    {
+      "epoch": 0.6513157894736842,
+      "grad_norm": 0.09075104025668673,
+      "learning_rate": 0.00019176086366117211,
+      "loss": 0.304,
+      "step": 495
+    },
+    {
+      "epoch": 0.6526315789473685,
+      "grad_norm": 0.0893794254125347,
+      "learning_rate": 0.0001916999015531221,
+      "loss": 0.2917,
+      "step": 496
+    },
+    {
+      "epoch": 0.6539473684210526,
+      "grad_norm": 0.089792074117038,
+      "learning_rate": 0.00019163872450676835,
+      "loss": 0.2812,
+      "step": 497
+    },
+    {
+      "epoch": 0.6552631578947369,
+      "grad_norm": 0.09280308668273284,
+      "learning_rate": 0.00019157733266550575,
+      "loss": 0.3024,
+      "step": 498
+    },
+    {
+      "epoch": 0.656578947368421,
+      "grad_norm": 0.0871128187479411,
+      "learning_rate": 0.00019151572617323253,
+      "loss": 0.29,
+      "step": 499
+    },
+    {
+      "epoch": 0.6578947368421053,
+      "grad_norm": 0.08870708451193264,
+      "learning_rate": 0.00019145390517435012,
+      "loss": 0.3034,
+      "step": 500
+    },
+    {
+      "epoch": 0.6592105263157895,
+      "grad_norm": 0.08720360293850979,
+      "learning_rate": 0.00019139186981376267,
+      "loss": 0.2844,
+      "step": 501
+    },
+    {
+      "epoch": 0.6605263157894737,
+      "grad_norm": 0.08832729646265089,
+      "learning_rate": 0.0001913296202368769,
+      "loss": 0.2999,
+      "step": 502
+    },
+    {
+      "epoch": 0.6618421052631579,
+      "grad_norm": 0.08817831223979289,
+      "learning_rate": 0.0001912671565896015,
+      "loss": 0.3034,
+      "step": 503
+    },
+    {
+      "epoch": 0.6631578947368421,
+      "grad_norm": 0.08919226453530668,
+      "learning_rate": 0.00019120447901834706,
+      "loss": 0.2865,
+      "step": 504
+    },
+    {
+      "epoch": 0.6644736842105263,
+      "grad_norm": 0.09256732471786443,
+      "learning_rate": 0.00019114158767002547,
+      "loss": 0.2887,
+      "step": 505
+    },
+    {
+      "epoch": 0.6657894736842105,
+      "grad_norm": 0.08807553839373092,
+      "learning_rate": 0.00019107848269204976,
+      "loss": 0.2864,
+      "step": 506
+    },
+    {
+      "epoch": 0.6671052631578948,
+      "grad_norm": 0.08590339271381157,
+      "learning_rate": 0.00019101516423233368,
+      "loss": 0.2878,
+      "step": 507
+    },
+    {
+      "epoch": 0.6684210526315789,
+      "grad_norm": 0.08651840397545302,
+      "learning_rate": 0.00019095163243929142,
+      "loss": 0.3005,
+      "step": 508
+    },
+    {
+      "epoch": 0.6697368421052632,
+      "grad_norm": 0.08955034533153189,
+      "learning_rate": 0.00019088788746183714,
+      "loss": 0.2906,
+      "step": 509
+    },
+    {
+      "epoch": 0.6710526315789473,
+      "grad_norm": 0.08824857512769824,
+      "learning_rate": 0.00019082392944938466,
+      "loss": 0.2826,
+      "step": 510
+    },
+    {
+      "epoch": 0.6723684210526316,
+      "grad_norm": 0.08981332868291347,
+      "learning_rate": 0.00019075975855184724,
+      "loss": 0.2791,
+      "step": 511
+    },
+    {
+      "epoch": 0.6736842105263158,
+      "grad_norm": 0.08872229343602653,
+      "learning_rate": 0.0001906953749196371,
+      "loss": 0.2941,
+      "step": 512
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 0.08604148465349523,
+      "learning_rate": 0.000190630778703665,
+      "loss": 0.2898,
+      "step": 513
+    },
+    {
+      "epoch": 0.6763157894736842,
+      "grad_norm": 0.09068748827778476,
+      "learning_rate": 0.00019056597005534013,
+      "loss": 0.3052,
+      "step": 514
+    },
+    {
+      "epoch": 0.6776315789473685,
+      "grad_norm": 0.09179119755745117,
+      "learning_rate": 0.00019050094912656952,
+      "loss": 0.3034,
+      "step": 515
+    },
+    {
+      "epoch": 0.6789473684210526,
+      "grad_norm": 0.08427846310926745,
+      "learning_rate": 0.00019043571606975777,
+      "loss": 0.2948,
+      "step": 516
+    },
+    {
+      "epoch": 0.6802631578947368,
+      "grad_norm": 0.08476965472606288,
+      "learning_rate": 0.00019037027103780668,
+      "loss": 0.2883,
+      "step": 517
+    },
+    {
+      "epoch": 0.6815789473684211,
+      "grad_norm": 0.08421580836098888,
+      "learning_rate": 0.00019030461418411497,
+      "loss": 0.2775,
+      "step": 518
+    },
+    {
+      "epoch": 0.6828947368421052,
+      "grad_norm": 0.08618030098765003,
+      "learning_rate": 0.00019023874566257784,
+      "loss": 0.2771,
+      "step": 519
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.08571261541013994,
+      "learning_rate": 0.00019017266562758659,
+      "loss": 0.289,
+      "step": 520
+    },
+    {
+      "epoch": 0.6855263157894737,
+      "grad_norm": 0.09462804464406739,
+      "learning_rate": 0.00019010637423402823,
+      "loss": 0.31,
+      "step": 521
+    },
+    {
+      "epoch": 0.6868421052631579,
+      "grad_norm": 0.09220006978220673,
+      "learning_rate": 0.00019003987163728535,
+      "loss": 0.2898,
+      "step": 522
+    },
+    {
+      "epoch": 0.6881578947368421,
+      "grad_norm": 0.08687665486527996,
+      "learning_rate": 0.00018997315799323548,
+      "loss": 0.2914,
+      "step": 523
+    },
+    {
+      "epoch": 0.6894736842105263,
+      "grad_norm": 0.08629627380926978,
+      "learning_rate": 0.00018990623345825083,
+      "loss": 0.2875,
+      "step": 524
+    },
+    {
+      "epoch": 0.6907894736842105,
+      "grad_norm": 0.08990734481501246,
+      "learning_rate": 0.0001898390981891979,
+      "loss": 0.2935,
+      "step": 525
+    },
+    {
+      "epoch": 0.6921052631578948,
+      "grad_norm": 0.09107845137785629,
+      "learning_rate": 0.00018977175234343723,
+      "loss": 0.293,
+      "step": 526
+    },
+    {
+      "epoch": 0.6934210526315789,
+      "grad_norm": 0.08746893451432518,
+      "learning_rate": 0.00018970419607882284,
+      "loss": 0.2834,
+      "step": 527
+    },
+    {
+      "epoch": 0.6947368421052632,
+      "grad_norm": 0.08510400660112896,
+      "learning_rate": 0.00018963642955370201,
+      "loss": 0.2836,
+      "step": 528
+    },
+    {
+      "epoch": 0.6960526315789474,
+      "grad_norm": 0.08902379266326126,
+      "learning_rate": 0.00018956845292691487,
+      "loss": 0.2918,
+      "step": 529
+    },
+    {
+      "epoch": 0.6973684210526315,
+      "grad_norm": 0.08716702459764675,
+      "learning_rate": 0.00018950026635779397,
+      "loss": 0.3018,
+      "step": 530
+    },
+    {
+      "epoch": 0.6986842105263158,
+      "grad_norm": 0.0860349723743795,
+      "learning_rate": 0.00018943187000616395,
+      "loss": 0.2862,
+      "step": 531
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08640233925578623,
+      "learning_rate": 0.00018936326403234125,
+      "loss": 0.281,
+      "step": 532
+    },
+    {
+      "epoch": 0.7013157894736842,
+      "grad_norm": 0.08958858699111341,
+      "learning_rate": 0.0001892944485971335,
+      "loss": 0.2835,
+      "step": 533
+    },
+    {
+      "epoch": 0.7026315789473684,
+      "grad_norm": 0.09155021097028584,
+      "learning_rate": 0.0001892254238618394,
+      "loss": 0.2871,
+      "step": 534
+    },
+    {
+      "epoch": 0.7039473684210527,
+      "grad_norm": 0.08704346353009886,
+      "learning_rate": 0.00018915618998824825,
+      "loss": 0.2773,
+      "step": 535
+    },
+    {
+      "epoch": 0.7052631578947368,
+      "grad_norm": 0.09429002348494249,
+      "learning_rate": 0.00018908674713863952,
+      "loss": 0.2995,
+      "step": 536
+    },
+    {
+      "epoch": 0.7065789473684211,
+      "grad_norm": 0.0872284448725193,
+      "learning_rate": 0.00018901709547578245,
+      "loss": 0.29,
+      "step": 537
+    },
+    {
+      "epoch": 0.7078947368421052,
+      "grad_norm": 0.08323277604666501,
+      "learning_rate": 0.00018894723516293583,
+      "loss": 0.2924,
+      "step": 538
+    },
+    {
+      "epoch": 0.7092105263157895,
+      "grad_norm": 0.09209955844526282,
+      "learning_rate": 0.00018887716636384745,
+      "loss": 0.3093,
+      "step": 539
+    },
+    {
+      "epoch": 0.7105263157894737,
+      "grad_norm": 0.0928458401326009,
+      "learning_rate": 0.00018880688924275378,
+      "loss": 0.2898,
+      "step": 540
+    },
+    {
+      "epoch": 0.7118421052631579,
+      "grad_norm": 0.08447656093725353,
+      "learning_rate": 0.00018873640396437958,
+      "loss": 0.2998,
+      "step": 541
+    },
+    {
+      "epoch": 0.7131578947368421,
+      "grad_norm": 0.08914403154111118,
+      "learning_rate": 0.00018866571069393753,
+      "loss": 0.2975,
+      "step": 542
+    },
+    {
+      "epoch": 0.7144736842105263,
+      "grad_norm": 0.08061793607147465,
+      "learning_rate": 0.0001885948095971278,
+      "loss": 0.2693,
+      "step": 543
+    },
+    {
+      "epoch": 0.7157894736842105,
+      "grad_norm": 0.08081926207565744,
+      "learning_rate": 0.0001885237008401378,
+      "loss": 0.2833,
+      "step": 544
+    },
+    {
+      "epoch": 0.7171052631578947,
+      "grad_norm": 0.08814474168945655,
+      "learning_rate": 0.00018845238458964155,
+      "loss": 0.2867,
+      "step": 545
+    },
+    {
+      "epoch": 0.718421052631579,
+      "grad_norm": 0.08581179650874697,
+      "learning_rate": 0.00018838086101279945,
+      "loss": 0.2966,
+      "step": 546
+    },
+    {
+      "epoch": 0.7197368421052631,
+      "grad_norm": 0.08824691850575504,
+      "learning_rate": 0.0001883091302772579,
+      "loss": 0.2855,
+      "step": 547
+    },
+    {
+      "epoch": 0.7210526315789474,
+      "grad_norm": 0.08188688143306905,
+      "learning_rate": 0.0001882371925511488,
+      "loss": 0.2913,
+      "step": 548
+    },
+    {
+      "epoch": 0.7223684210526315,
+      "grad_norm": 0.08584104136037612,
+      "learning_rate": 0.00018816504800308934,
+      "loss": 0.2907,
+      "step": 549
+    },
+    {
+      "epoch": 0.7236842105263158,
+      "grad_norm": 0.08410484967989623,
+      "learning_rate": 0.00018809269680218136,
+      "loss": 0.2745,
+      "step": 550
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 0.08587157095386123,
+      "learning_rate": 0.00018802013911801112,
+      "loss": 0.2881,
+      "step": 551
+    },
+    {
+      "epoch": 0.7263157894736842,
+      "grad_norm": 0.09288984345534493,
+      "learning_rate": 0.0001879473751206489,
+      "loss": 0.3067,
+      "step": 552
+    },
+    {
+      "epoch": 0.7276315789473684,
+      "grad_norm": 0.08697284865649188,
+      "learning_rate": 0.00018787440498064856,
+      "loss": 0.2919,
+      "step": 553
+    },
+    {
+      "epoch": 0.7289473684210527,
+      "grad_norm": 0.08793299552682735,
+      "learning_rate": 0.00018780122886904709,
+      "loss": 0.3029,
+      "step": 554
+    },
+    {
+      "epoch": 0.7302631578947368,
+      "grad_norm": 0.08461397738242368,
+      "learning_rate": 0.0001877278469573643,
+      "loss": 0.2977,
+      "step": 555
+    },
+    {
+      "epoch": 0.7315789473684211,
+      "grad_norm": 0.0874714643672974,
+      "learning_rate": 0.00018765425941760238,
+      "loss": 0.2928,
+      "step": 556
+    },
+    {
+      "epoch": 0.7328947368421053,
+      "grad_norm": 0.0873090646029582,
+      "learning_rate": 0.0001875804664222455,
+      "loss": 0.3066,
+      "step": 557
+    },
+    {
+      "epoch": 0.7342105263157894,
+      "grad_norm": 0.08379654737914938,
+      "learning_rate": 0.00018750646814425938,
+      "loss": 0.2977,
+      "step": 558
+    },
+    {
+      "epoch": 0.7355263157894737,
+      "grad_norm": 0.09314788862619253,
+      "learning_rate": 0.00018743226475709094,
+      "loss": 0.2963,
+      "step": 559
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.08588334887643348,
+      "learning_rate": 0.00018735785643466784,
+      "loss": 0.2878,
+      "step": 560
+    },
+    {
+      "epoch": 0.7381578947368421,
+      "grad_norm": 0.08163329639613695,
+      "learning_rate": 0.00018728324335139814,
+      "loss": 0.2991,
+      "step": 561
+    },
+    {
+      "epoch": 0.7394736842105263,
+      "grad_norm": 0.08439090485689141,
+      "learning_rate": 0.00018720842568216978,
+      "loss": 0.274,
+      "step": 562
+    },
+    {
+      "epoch": 0.7407894736842106,
+      "grad_norm": 0.09031678503562862,
+      "learning_rate": 0.0001871334036023503,
+      "loss": 0.2942,
+      "step": 563
+    },
+    {
+      "epoch": 0.7421052631578947,
+      "grad_norm": 0.0812973217567514,
+      "learning_rate": 0.00018705817728778624,
+      "loss": 0.2725,
+      "step": 564
+    },
+    {
+      "epoch": 0.743421052631579,
+      "grad_norm": 0.08201157379247533,
+      "learning_rate": 0.00018698274691480302,
+      "loss": 0.2712,
+      "step": 565
+    },
+    {
+      "epoch": 0.7447368421052631,
+      "grad_norm": 0.08733882187748677,
+      "learning_rate": 0.00018690711266020426,
+      "loss": 0.2841,
+      "step": 566
+    },
+    {
+      "epoch": 0.7460526315789474,
+      "grad_norm": 0.08741424395823033,
+      "learning_rate": 0.0001868312747012715,
+      "loss": 0.2763,
+      "step": 567
+    },
+    {
+      "epoch": 0.7473684210526316,
+      "grad_norm": 0.08660511543001677,
+      "learning_rate": 0.00018675523321576371,
+      "loss": 0.2847,
+      "step": 568
+    },
+    {
+      "epoch": 0.7486842105263158,
+      "grad_norm": 0.08668938841950703,
+      "learning_rate": 0.00018667898838191694,
+      "loss": 0.2821,
+      "step": 569
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08397772374383645,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 0.2919,
+      "step": 570
+    },
+    {
+      "epoch": 0.7513157894736842,
+      "grad_norm": 0.0890009207591882,
+      "learning_rate": 0.0001865258893845334,
+      "loss": 0.2935,
+      "step": 571
+    },
+    {
+      "epoch": 0.7526315789473684,
+      "grad_norm": 0.08615071748725678,
+      "learning_rate": 0.00018644903557985025,
+      "loss": 0.2862,
+      "step": 572
+    },
+    {
+      "epoch": 0.7539473684210526,
+      "grad_norm": 0.08035895214007202,
+      "learning_rate": 0.00018637197914453445,
+      "loss": 0.2915,
+      "step": 573
+    },
+    {
+      "epoch": 0.7552631578947369,
+      "grad_norm": 0.08394144603894776,
+      "learning_rate": 0.000186294720259201,
+      "loss": 0.2905,
+      "step": 574
+    },
+    {
+      "epoch": 0.756578947368421,
+      "grad_norm": 0.08681582101826608,
+      "learning_rate": 0.0001862172591049395,
+      "loss": 0.2927,
+      "step": 575
+    },
+    {
+      "epoch": 0.7578947368421053,
+      "grad_norm": 0.08027258031476615,
+      "learning_rate": 0.00018613959586331362,
+      "loss": 0.2717,
+      "step": 576
+    },
+    {
+      "epoch": 0.7592105263157894,
+      "grad_norm": 0.0832571328383153,
+      "learning_rate": 0.0001860617307163606,
+      "loss": 0.2877,
+      "step": 577
+    },
+    {
+      "epoch": 0.7605263157894737,
+      "grad_norm": 0.0855055635944409,
+      "learning_rate": 0.0001859836638465911,
+      "loss": 0.2879,
+      "step": 578
+    },
+    {
+      "epoch": 0.7618421052631579,
+      "grad_norm": 0.08455171659738935,
+      "learning_rate": 0.00018590539543698854,
+      "loss": 0.2946,
+      "step": 579
+    },
+    {
+      "epoch": 0.7631578947368421,
+      "grad_norm": 0.08740787324026139,
+      "learning_rate": 0.00018582692567100867,
+      "loss": 0.2835,
+      "step": 580
+    },
+    {
+      "epoch": 0.7644736842105263,
+      "grad_norm": 0.08908515189955798,
+      "learning_rate": 0.00018574825473257925,
+      "loss": 0.2917,
+      "step": 581
+    },
+    {
+      "epoch": 0.7657894736842106,
+      "grad_norm": 0.0856923697970573,
+      "learning_rate": 0.00018566938280609966,
+      "loss": 0.2794,
+      "step": 582
+    },
+    {
+      "epoch": 0.7671052631578947,
+      "grad_norm": 0.08400580488406503,
+      "learning_rate": 0.00018559031007644024,
+      "loss": 0.2764,
+      "step": 583
+    },
+    {
+      "epoch": 0.7684210526315789,
+      "grad_norm": 0.08599213131358657,
+      "learning_rate": 0.00018551103672894206,
+      "loss": 0.2902,
+      "step": 584
+    },
+    {
+      "epoch": 0.7697368421052632,
+      "grad_norm": 0.08612768497888425,
+      "learning_rate": 0.0001854315629494165,
+      "loss": 0.2913,
+      "step": 585
+    },
+    {
+      "epoch": 0.7710526315789473,
+      "grad_norm": 0.08848005818883788,
+      "learning_rate": 0.0001853518889241446,
+      "loss": 0.2913,
+      "step": 586
+    },
+    {
+      "epoch": 0.7723684210526316,
+      "grad_norm": 0.08443059614838172,
+      "learning_rate": 0.0001852720148398769,
+      "loss": 0.283,
+      "step": 587
+    },
+    {
+      "epoch": 0.7736842105263158,
+      "grad_norm": 0.08864582533090447,
+      "learning_rate": 0.00018519194088383273,
+      "loss": 0.2888,
+      "step": 588
+    },
+    {
+      "epoch": 0.775,
+      "grad_norm": 0.08840947193889603,
+      "learning_rate": 0.00018511166724369997,
+      "loss": 0.2931,
+      "step": 589
+    },
+    {
+      "epoch": 0.7763157894736842,
+      "grad_norm": 0.08363366619678043,
+      "learning_rate": 0.0001850311941076346,
+      "loss": 0.2782,
+      "step": 590
+    },
+    {
+      "epoch": 0.7776315789473685,
+      "grad_norm": 0.08538148462925309,
+      "learning_rate": 0.00018495052166426015,
+      "loss": 0.2856,
+      "step": 591
+    },
+    {
+      "epoch": 0.7789473684210526,
+      "grad_norm": 0.08343535715617335,
+      "learning_rate": 0.00018486965010266725,
+      "loss": 0.2847,
+      "step": 592
+    },
+    {
+      "epoch": 0.7802631578947369,
+      "grad_norm": 0.08230524033132103,
+      "learning_rate": 0.00018478857961241337,
+      "loss": 0.2878,
+      "step": 593
+    },
+    {
+      "epoch": 0.781578947368421,
+      "grad_norm": 0.08479401897743853,
+      "learning_rate": 0.0001847073103835222,
+      "loss": 0.2873,
+      "step": 594
+    },
+    {
+      "epoch": 0.7828947368421053,
+      "grad_norm": 0.08524719297922047,
+      "learning_rate": 0.00018462584260648323,
+      "loss": 0.2845,
+      "step": 595
+    },
+    {
+      "epoch": 0.7842105263157895,
+      "grad_norm": 0.08448036993692473,
+      "learning_rate": 0.0001845441764722514,
+      "loss": 0.288,
+      "step": 596
+    },
+    {
+      "epoch": 0.7855263157894737,
+      "grad_norm": 0.08740687637602888,
+      "learning_rate": 0.0001844623121722465,
+      "loss": 0.2753,
+      "step": 597
+    },
+    {
+      "epoch": 0.7868421052631579,
+      "grad_norm": 0.08303399012001453,
+      "learning_rate": 0.0001843802498983529,
+      "loss": 0.2875,
+      "step": 598
+    },
+    {
+      "epoch": 0.7881578947368421,
+      "grad_norm": 0.08523557992775858,
+      "learning_rate": 0.00018429798984291896,
+      "loss": 0.2843,
+      "step": 599
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.0882338916613198,
+      "learning_rate": 0.00018421553219875658,
+      "loss": 0.3044,
+      "step": 600
+    },
+    {
+      "epoch": 0.7907894736842105,
+      "grad_norm": 0.08386004069859475,
+      "learning_rate": 0.00018413287715914089,
+      "loss": 0.2744,
+      "step": 601
+    },
+    {
+      "epoch": 0.7921052631578948,
+      "grad_norm": 0.0860906320981426,
+      "learning_rate": 0.00018405002491780968,
+      "loss": 0.2904,
+      "step": 602
+    },
+    {
+      "epoch": 0.7934210526315789,
+      "grad_norm": 0.08412547381525294,
+      "learning_rate": 0.00018396697566896286,
+      "loss": 0.2944,
+      "step": 603
+    },
+    {
+      "epoch": 0.7947368421052632,
+      "grad_norm": 0.08160487708739002,
+      "learning_rate": 0.00018388372960726228,
+      "loss": 0.2894,
+      "step": 604
+    },
+    {
+      "epoch": 0.7960526315789473,
+      "grad_norm": 0.07789675625414566,
+      "learning_rate": 0.00018380028692783096,
+      "loss": 0.289,
+      "step": 605
+    },
+    {
+      "epoch": 0.7973684210526316,
+      "grad_norm": 0.08326388068355157,
+      "learning_rate": 0.00018371664782625287,
+      "loss": 0.2757,
+      "step": 606
+    },
+    {
+      "epoch": 0.7986842105263158,
+      "grad_norm": 0.0813892959155161,
+      "learning_rate": 0.00018363281249857233,
+      "loss": 0.2805,
+      "step": 607
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08338896150212435,
+      "learning_rate": 0.00018354878114129367,
+      "loss": 0.2761,
+      "step": 608
+    },
+    {
+      "epoch": 0.8013157894736842,
+      "grad_norm": 0.08472324714205545,
+      "learning_rate": 0.00018346455395138058,
+      "loss": 0.2814,
+      "step": 609
+    },
+    {
+      "epoch": 0.8026315789473685,
+      "grad_norm": 0.08513461063729397,
+      "learning_rate": 0.00018338013112625587,
+      "loss": 0.2952,
+      "step": 610
+    },
+    {
+      "epoch": 0.8039473684210526,
+      "grad_norm": 0.08100863304109138,
+      "learning_rate": 0.00018329551286380087,
+      "loss": 0.2756,
+      "step": 611
+    },
+    {
+      "epoch": 0.8052631578947368,
+      "grad_norm": 0.0817298921819726,
+      "learning_rate": 0.00018321069936235503,
+      "loss": 0.2788,
+      "step": 612
+    },
+    {
+      "epoch": 0.8065789473684211,
+      "grad_norm": 0.08139556434473312,
+      "learning_rate": 0.00018312569082071535,
+      "loss": 0.2855,
+      "step": 613
+    },
+    {
+      "epoch": 0.8078947368421052,
+      "grad_norm": 0.08426481750684653,
+      "learning_rate": 0.0001830404874381361,
+      "loss": 0.2932,
+      "step": 614
+    },
+    {
+      "epoch": 0.8092105263157895,
+      "grad_norm": 0.08338650300983878,
+      "learning_rate": 0.00018295508941432815,
+      "loss": 0.291,
+      "step": 615
+    },
+    {
+      "epoch": 0.8105263157894737,
+      "grad_norm": 0.0841355163206272,
+      "learning_rate": 0.00018286949694945866,
+      "loss": 0.2905,
+      "step": 616
+    },
+    {
+      "epoch": 0.8118421052631579,
+      "grad_norm": 0.08411824810347067,
+      "learning_rate": 0.0001827837102441505,
+      "loss": 0.2992,
+      "step": 617
+    },
+    {
+      "epoch": 0.8131578947368421,
+      "grad_norm": 0.08090022262513084,
+      "learning_rate": 0.00018269772949948182,
+      "loss": 0.2906,
+      "step": 618
+    },
+    {
+      "epoch": 0.8144736842105263,
+      "grad_norm": 0.0788680440518934,
+      "learning_rate": 0.00018261155491698568,
+      "loss": 0.2939,
+      "step": 619
+    },
+    {
+      "epoch": 0.8157894736842105,
+      "grad_norm": 0.07833398942854092,
+      "learning_rate": 0.00018252518669864936,
+      "loss": 0.2873,
+      "step": 620
+    },
+    {
+      "epoch": 0.8171052631578948,
+      "grad_norm": 0.08646064241825135,
+      "learning_rate": 0.00018243862504691407,
+      "loss": 0.2831,
+      "step": 621
+    },
+    {
+      "epoch": 0.8184210526315789,
+      "grad_norm": 0.07939171423489742,
+      "learning_rate": 0.00018235187016467442,
+      "loss": 0.2831,
+      "step": 622
+    },
+    {
+      "epoch": 0.8197368421052632,
+      "grad_norm": 0.08346477115007297,
+      "learning_rate": 0.0001822649222552779,
+      "loss": 0.2832,
+      "step": 623
+    },
+    {
+      "epoch": 0.8210526315789474,
+      "grad_norm": 0.08777211185850274,
+      "learning_rate": 0.0001821777815225245,
+      "loss": 0.2946,
+      "step": 624
+    },
+    {
+      "epoch": 0.8223684210526315,
+      "grad_norm": 0.08176020142698397,
+      "learning_rate": 0.00018209044817066617,
+      "loss": 0.2808,
+      "step": 625
+    },
+    {
+      "epoch": 0.8236842105263158,
+      "grad_norm": 0.083353664452084,
+      "learning_rate": 0.00018200292240440623,
+      "loss": 0.2964,
+      "step": 626
+    },
+    {
+      "epoch": 0.825,
+      "grad_norm": 0.08320257861919898,
+      "learning_rate": 0.0001819152044288992,
+      "loss": 0.2687,
+      "step": 627
+    },
+    {
+      "epoch": 0.8263157894736842,
+      "grad_norm": 0.08315187924191537,
+      "learning_rate": 0.00018182729444974992,
+      "loss": 0.2911,
+      "step": 628
+    },
+    {
+      "epoch": 0.8276315789473684,
+      "grad_norm": 0.08465028554778517,
+      "learning_rate": 0.00018173919267301344,
+      "loss": 0.2849,
+      "step": 629
+    },
+    {
+      "epoch": 0.8289473684210527,
+      "grad_norm": 0.0808599203291463,
+      "learning_rate": 0.0001816508993051943,
+      "loss": 0.2842,
+      "step": 630
+    },
+    {
+      "epoch": 0.8302631578947368,
+      "grad_norm": 0.08319880407928407,
+      "learning_rate": 0.0001815624145532461,
+      "loss": 0.2878,
+      "step": 631
+    },
+    {
+      "epoch": 0.8315789473684211,
+      "grad_norm": 0.08454873694804513,
+      "learning_rate": 0.00018147373862457107,
+      "loss": 0.2977,
+      "step": 632
+    },
+    {
+      "epoch": 0.8328947368421052,
+      "grad_norm": 0.08522249341627423,
+      "learning_rate": 0.0001813848717270195,
+      "loss": 0.2962,
+      "step": 633
+    },
+    {
+      "epoch": 0.8342105263157895,
+      "grad_norm": 0.08804872368012215,
+      "learning_rate": 0.00018129581406888936,
+      "loss": 0.283,
+      "step": 634
+    },
+    {
+      "epoch": 0.8355263157894737,
+      "grad_norm": 0.08019331668231523,
+      "learning_rate": 0.00018120656585892572,
+      "loss": 0.2856,
+      "step": 635
+    },
+    {
+      "epoch": 0.8368421052631579,
+      "grad_norm": 0.08839395034159263,
+      "learning_rate": 0.00018111712730632022,
+      "loss": 0.3064,
+      "step": 636
+    },
+    {
+      "epoch": 0.8381578947368421,
+      "grad_norm": 0.08267930670342614,
+      "learning_rate": 0.00018102749862071083,
+      "loss": 0.2884,
+      "step": 637
+    },
+    {
+      "epoch": 0.8394736842105263,
+      "grad_norm": 0.08334284753944939,
+      "learning_rate": 0.00018093768001218094,
+      "loss": 0.2882,
+      "step": 638
+    },
+    {
+      "epoch": 0.8407894736842105,
+      "grad_norm": 0.08293903985279776,
+      "learning_rate": 0.00018084767169125932,
+      "loss": 0.2855,
+      "step": 639
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.08224010769882795,
+      "learning_rate": 0.0001807574738689193,
+      "loss": 0.3051,
+      "step": 640
+    },
+    {
+      "epoch": 0.843421052631579,
+      "grad_norm": 0.07771382955899404,
+      "learning_rate": 0.00018066708675657837,
+      "loss": 0.255,
+      "step": 641
+    },
+    {
+      "epoch": 0.8447368421052631,
+      "grad_norm": 0.08276207696773076,
+      "learning_rate": 0.00018057651056609784,
+      "loss": 0.2851,
+      "step": 642
+    },
+    {
+      "epoch": 0.8460526315789474,
+      "grad_norm": 0.08138572227367649,
+      "learning_rate": 0.000180485745509782,
+      "loss": 0.2726,
+      "step": 643
+    },
+    {
+      "epoch": 0.8473684210526315,
+      "grad_norm": 0.07991337682516308,
+      "learning_rate": 0.000180394791800378,
+      "loss": 0.2795,
+      "step": 644
+    },
+    {
+      "epoch": 0.8486842105263158,
+      "grad_norm": 0.08368290547552978,
+      "learning_rate": 0.0001803036496510752,
+      "loss": 0.2876,
+      "step": 645
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08602046286102522,
+      "learning_rate": 0.0001802123192755044,
+      "loss": 0.2899,
+      "step": 646
+    },
+    {
+      "epoch": 0.8513157894736842,
+      "grad_norm": 0.08247602682736774,
+      "learning_rate": 0.00018012080088773786,
+      "loss": 0.289,
+      "step": 647
+    },
+    {
+      "epoch": 0.8526315789473684,
+      "grad_norm": 0.08076494362770659,
+      "learning_rate": 0.00018002909470228842,
+      "loss": 0.2953,
+      "step": 648
+    },
+    {
+      "epoch": 0.8539473684210527,
+      "grad_norm": 0.08213269492727011,
+      "learning_rate": 0.0001799372009341091,
+      "loss": 0.2834,
+      "step": 649
+    },
+    {
+      "epoch": 0.8552631578947368,
+      "grad_norm": 0.07906596825371759,
+      "learning_rate": 0.00017984511979859263,
+      "loss": 0.2884,
+      "step": 650
+    },
+    {
+      "epoch": 0.8565789473684211,
+      "grad_norm": 0.08280102631376626,
+      "learning_rate": 0.0001797528515115709,
+      "loss": 0.2898,
+      "step": 651
+    },
+    {
+      "epoch": 0.8578947368421053,
+      "grad_norm": 0.08195845362740756,
+      "learning_rate": 0.00017966039628931446,
+      "loss": 0.2811,
+      "step": 652
+    },
+    {
+      "epoch": 0.8592105263157894,
+      "grad_norm": 0.08275884521115351,
+      "learning_rate": 0.00017956775434853201,
+      "loss": 0.2809,
+      "step": 653
+    },
+    {
+      "epoch": 0.8605263157894737,
+      "grad_norm": 0.0823561522762983,
+      "learning_rate": 0.00017947492590637,
+      "loss": 0.2733,
+      "step": 654
+    },
+    {
+      "epoch": 0.8618421052631579,
+      "grad_norm": 0.08408575804402002,
+      "learning_rate": 0.00017938191118041185,
+      "loss": 0.2921,
+      "step": 655
+    },
+    {
+      "epoch": 0.8631578947368421,
+      "grad_norm": 0.08263210898936545,
+      "learning_rate": 0.00017928871038867784,
+      "loss": 0.2896,
+      "step": 656
+    },
+    {
+      "epoch": 0.8644736842105263,
+      "grad_norm": 0.08282963133814529,
+      "learning_rate": 0.00017919532374962416,
+      "loss": 0.267,
+      "step": 657
+    },
+    {
+      "epoch": 0.8657894736842106,
+      "grad_norm": 0.08456674605493757,
+      "learning_rate": 0.00017910175148214274,
+      "loss": 0.2846,
+      "step": 658
+    },
+    {
+      "epoch": 0.8671052631578947,
+      "grad_norm": 0.08295032078815928,
+      "learning_rate": 0.00017900799380556065,
+      "loss": 0.2957,
+      "step": 659
+    },
+    {
+      "epoch": 0.868421052631579,
+      "grad_norm": 0.08407937334549097,
+      "learning_rate": 0.00017891405093963938,
+      "loss": 0.2801,
+      "step": 660
+    },
+    {
+      "epoch": 0.8697368421052631,
+      "grad_norm": 0.08152847255098901,
+      "learning_rate": 0.00017881992310457461,
+      "loss": 0.287,
+      "step": 661
+    },
+    {
+      "epoch": 0.8710526315789474,
+      "grad_norm": 0.07846383233153922,
+      "learning_rate": 0.00017872561052099562,
+      "loss": 0.2808,
+      "step": 662
+    },
+    {
+      "epoch": 0.8723684210526316,
+      "grad_norm": 0.08455839714844739,
+      "learning_rate": 0.00017863111340996458,
+      "loss": 0.2887,
+      "step": 663
+    },
+    {
+      "epoch": 0.8736842105263158,
+      "grad_norm": 0.08214963388362358,
+      "learning_rate": 0.00017853643199297633,
+      "loss": 0.2826,
+      "step": 664
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.08555672825696439,
+      "learning_rate": 0.00017844156649195759,
+      "loss": 0.2933,
+      "step": 665
+    },
+    {
+      "epoch": 0.8763157894736842,
+      "grad_norm": 0.08460096862590989,
+      "learning_rate": 0.00017834651712926662,
+      "loss": 0.2826,
+      "step": 666
+    },
+    {
+      "epoch": 0.8776315789473684,
+      "grad_norm": 0.08252349078451936,
+      "learning_rate": 0.00017825128412769266,
+      "loss": 0.2958,
+      "step": 667
+    },
+    {
+      "epoch": 0.8789473684210526,
+      "grad_norm": 0.08712916060067398,
+      "learning_rate": 0.00017815586771045535,
+      "loss": 0.3058,
+      "step": 668
+    },
+    {
+      "epoch": 0.8802631578947369,
+      "grad_norm": 0.08180591687568112,
+      "learning_rate": 0.00017806026810120423,
+      "loss": 0.2895,
+      "step": 669
+    },
+    {
+      "epoch": 0.881578947368421,
+      "grad_norm": 0.07876973529889683,
+      "learning_rate": 0.00017796448552401825,
+      "loss": 0.286,
+      "step": 670
+    },
+    {
+      "epoch": 0.8828947368421053,
+      "grad_norm": 0.07992980593985875,
+      "learning_rate": 0.00017786852020340525,
+      "loss": 0.2861,
+      "step": 671
+    },
+    {
+      "epoch": 0.8842105263157894,
+      "grad_norm": 0.08596031959873009,
+      "learning_rate": 0.0001777723723643014,
+      "loss": 0.2956,
+      "step": 672
+    },
+    {
+      "epoch": 0.8855263157894737,
+      "grad_norm": 0.08250501195390463,
+      "learning_rate": 0.00017767604223207064,
+      "loss": 0.2925,
+      "step": 673
+    },
+    {
+      "epoch": 0.8868421052631579,
+      "grad_norm": 0.08065929962541825,
+      "learning_rate": 0.00017757953003250422,
+      "loss": 0.2751,
+      "step": 674
+    },
+    {
+      "epoch": 0.8881578947368421,
+      "grad_norm": 0.079685175906907,
+      "learning_rate": 0.00017748283599182014,
+      "loss": 0.2713,
+      "step": 675
+    },
+    {
+      "epoch": 0.8894736842105263,
+      "grad_norm": 0.08275584920700671,
+      "learning_rate": 0.0001773859603366626,
+      "loss": 0.2805,
+      "step": 676
+    },
+    {
+      "epoch": 0.8907894736842106,
+      "grad_norm": 0.08237909937314071,
+      "learning_rate": 0.00017728890329410157,
+      "loss": 0.2769,
+      "step": 677
+    },
+    {
+      "epoch": 0.8921052631578947,
+      "grad_norm": 0.08027830306430328,
+      "learning_rate": 0.0001771916650916321,
+      "loss": 0.2815,
+      "step": 678
+    },
+    {
+      "epoch": 0.8934210526315789,
+      "grad_norm": 0.07965207648580828,
+      "learning_rate": 0.00017709424595717388,
+      "loss": 0.2767,
+      "step": 679
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.08200170949634054,
+      "learning_rate": 0.00017699664611907072,
+      "loss": 0.2671,
+      "step": 680
+    },
+    {
+      "epoch": 0.8960526315789473,
+      "grad_norm": 0.0837945547867222,
+      "learning_rate": 0.00017689886580608998,
+      "loss": 0.2846,
+      "step": 681
+    },
+    {
+      "epoch": 0.8973684210526316,
+      "grad_norm": 0.08287411520644314,
+      "learning_rate": 0.00017680090524742204,
+      "loss": 0.2797,
+      "step": 682
+    },
+    {
+      "epoch": 0.8986842105263158,
+      "grad_norm": 0.08346778301448879,
+      "learning_rate": 0.0001767027646726797,
+      "loss": 0.2851,
+      "step": 683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08078331578360252,
+      "learning_rate": 0.0001766044443118978,
+      "loss": 0.2758,
+      "step": 684
+    },
+    {
+      "epoch": 0.9013157894736842,
+      "grad_norm": 0.0823050633372224,
+      "learning_rate": 0.0001765059443955326,
+      "loss": 0.2896,
+      "step": 685
+    },
+    {
+      "epoch": 0.9026315789473685,
+      "grad_norm": 0.08085978291636725,
+      "learning_rate": 0.00017640726515446103,
+      "loss": 0.2752,
+      "step": 686
+    },
+    {
+      "epoch": 0.9039473684210526,
+      "grad_norm": 0.08633725156118303,
+      "learning_rate": 0.00017630840681998066,
+      "loss": 0.2885,
+      "step": 687
+    },
+    {
+      "epoch": 0.9052631578947369,
+      "grad_norm": 0.08134705054643286,
+      "learning_rate": 0.00017620936962380856,
+      "loss": 0.2741,
+      "step": 688
+    },
+    {
+      "epoch": 0.906578947368421,
+      "grad_norm": 0.08218959378945977,
+      "learning_rate": 0.0001761101537980812,
+      "loss": 0.2932,
+      "step": 689
+    },
+    {
+      "epoch": 0.9078947368421053,
+      "grad_norm": 0.08050157591340845,
+      "learning_rate": 0.00017601075957535364,
+      "loss": 0.2801,
+      "step": 690
+    },
+    {
+      "epoch": 0.9092105263157895,
+      "grad_norm": 0.07842237436882951,
+      "learning_rate": 0.00017591118718859923,
+      "loss": 0.2672,
+      "step": 691
+    },
+    {
+      "epoch": 0.9105263157894737,
+      "grad_norm": 0.08141690192884726,
+      "learning_rate": 0.00017581143687120875,
+      "loss": 0.2875,
+      "step": 692
+    },
+    {
+      "epoch": 0.9118421052631579,
+      "grad_norm": 0.0841178887567793,
+      "learning_rate": 0.00017571150885699023,
+      "loss": 0.288,
+      "step": 693
+    },
+    {
+      "epoch": 0.9131578947368421,
+      "grad_norm": 0.08199498281628458,
+      "learning_rate": 0.00017561140338016802,
+      "loss": 0.2908,
+      "step": 694
+    },
+    {
+      "epoch": 0.9144736842105263,
+      "grad_norm": 0.07861915918731534,
+      "learning_rate": 0.00017551112067538255,
+      "loss": 0.2714,
+      "step": 695
+    },
+    {
+      "epoch": 0.9157894736842105,
+      "grad_norm": 0.0831092654603713,
+      "learning_rate": 0.00017541066097768963,
+      "loss": 0.2829,
+      "step": 696
+    },
+    {
+      "epoch": 0.9171052631578948,
+      "grad_norm": 0.08057002540174062,
+      "learning_rate": 0.00017531002452255993,
+      "loss": 0.2938,
+      "step": 697
+    },
+    {
+      "epoch": 0.9184210526315789,
+      "grad_norm": 0.08109895602300159,
+      "learning_rate": 0.00017520921154587843,
+      "loss": 0.2913,
+      "step": 698
+    },
+    {
+      "epoch": 0.9197368421052632,
+      "grad_norm": 0.08102699248656946,
+      "learning_rate": 0.00017510822228394385,
+      "loss": 0.2725,
+      "step": 699
+    },
+    {
+      "epoch": 0.9210526315789473,
+      "grad_norm": 0.07787423365882763,
+      "learning_rate": 0.0001750070569734681,
+      "loss": 0.2741,
+      "step": 700
+    },
+    {
+      "epoch": 0.9223684210526316,
+      "grad_norm": 0.07824132469372493,
+      "learning_rate": 0.00017490571585157576,
+      "loss": 0.2723,
+      "step": 701
+    },
+    {
+      "epoch": 0.9236842105263158,
+      "grad_norm": 0.08382351085436456,
+      "learning_rate": 0.00017480419915580356,
+      "loss": 0.284,
+      "step": 702
+    },
+    {
+      "epoch": 0.925,
+      "grad_norm": 0.0807625674700622,
+      "learning_rate": 0.0001747025071240996,
+      "loss": 0.2749,
+      "step": 703
+    },
+    {
+      "epoch": 0.9263157894736842,
+      "grad_norm": 0.0816515881967252,
+      "learning_rate": 0.00017460063999482316,
+      "loss": 0.2789,
+      "step": 704
+    },
+    {
+      "epoch": 0.9276315789473685,
+      "grad_norm": 0.07777785520204356,
+      "learning_rate": 0.00017449859800674371,
+      "loss": 0.2713,
+      "step": 705
+    },
+    {
+      "epoch": 0.9289473684210526,
+      "grad_norm": 0.08076226646853048,
+      "learning_rate": 0.0001743963813990408,
+      "loss": 0.2953,
+      "step": 706
+    },
+    {
+      "epoch": 0.9302631578947368,
+      "grad_norm": 0.07973740349278101,
+      "learning_rate": 0.00017429399041130313,
+      "loss": 0.2662,
+      "step": 707
+    },
+    {
+      "epoch": 0.9315789473684211,
+      "grad_norm": 0.08456169813288093,
+      "learning_rate": 0.00017419142528352817,
+      "loss": 0.2859,
+      "step": 708
+    },
+    {
+      "epoch": 0.9328947368421052,
+      "grad_norm": 0.08035481674307465,
+      "learning_rate": 0.0001740886862561216,
+      "loss": 0.2809,
+      "step": 709
+    },
+    {
+      "epoch": 0.9342105263157895,
+      "grad_norm": 0.08259040712172755,
+      "learning_rate": 0.00017398577356989665,
+      "loss": 0.3024,
+      "step": 710
+    },
+    {
+      "epoch": 0.9355263157894737,
+      "grad_norm": 0.08153299418934971,
+      "learning_rate": 0.0001738826874660737,
+      "loss": 0.2871,
+      "step": 711
+    },
+    {
+      "epoch": 0.9368421052631579,
+      "grad_norm": 0.08013584006531711,
+      "learning_rate": 0.00017377942818627942,
+      "loss": 0.279,
+      "step": 712
+    },
+    {
+      "epoch": 0.9381578947368421,
+      "grad_norm": 0.08047700820955582,
+      "learning_rate": 0.00017367599597254655,
+      "loss": 0.2801,
+      "step": 713
+    },
+    {
+      "epoch": 0.9394736842105263,
+      "grad_norm": 0.07885052618166923,
+      "learning_rate": 0.00017357239106731317,
+      "loss": 0.2899,
+      "step": 714
+    },
+    {
+      "epoch": 0.9407894736842105,
+      "grad_norm": 0.07929961550190592,
+      "learning_rate": 0.0001734686137134221,
+      "loss": 0.2796,
+      "step": 715
+    },
+    {
+      "epoch": 0.9421052631578948,
+      "grad_norm": 0.08339513330893553,
+      "learning_rate": 0.00017336466415412028,
+      "loss": 0.2742,
+      "step": 716
+    },
+    {
+      "epoch": 0.9434210526315789,
+      "grad_norm": 0.07669794587650931,
+      "learning_rate": 0.00017326054263305847,
+      "loss": 0.2718,
+      "step": 717
+    },
+    {
+      "epoch": 0.9447368421052632,
+      "grad_norm": 0.0799990017084231,
+      "learning_rate": 0.00017315624939429037,
+      "loss": 0.287,
+      "step": 718
+    },
+    {
+      "epoch": 0.9460526315789474,
+      "grad_norm": 0.07817929739263474,
+      "learning_rate": 0.0001730517846822722,
+      "loss": 0.2883,
+      "step": 719
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.0795663214179471,
+      "learning_rate": 0.0001729471487418621,
+      "loss": 0.2693,
+      "step": 720
+    },
+    {
+      "epoch": 0.9486842105263158,
+      "grad_norm": 0.0837605598262032,
+      "learning_rate": 0.00017284234181831956,
+      "loss": 0.2771,
+      "step": 721
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08193295646472974,
+      "learning_rate": 0.00017273736415730488,
+      "loss": 0.3021,
+      "step": 722
+    },
+    {
+      "epoch": 0.9513157894736842,
+      "grad_norm": 0.08213211517441393,
+      "learning_rate": 0.00017263221600487852,
+      "loss": 0.2696,
+      "step": 723
+    },
+    {
+      "epoch": 0.9526315789473684,
+      "grad_norm": 0.08608126209062855,
+      "learning_rate": 0.0001725268976075005,
+      "loss": 0.2891,
+      "step": 724
+    },
+    {
+      "epoch": 0.9539473684210527,
+      "grad_norm": 0.07916611709429071,
+      "learning_rate": 0.00017242140921203003,
+      "loss": 0.2833,
+      "step": 725
+    },
+    {
+      "epoch": 0.9552631578947368,
+      "grad_norm": 0.08221381743802868,
+      "learning_rate": 0.00017231575106572467,
+      "loss": 0.2953,
+      "step": 726
+    },
+    {
+      "epoch": 0.9565789473684211,
+      "grad_norm": 0.07838487546103713,
+      "learning_rate": 0.0001722099234162399,
+      "loss": 0.2854,
+      "step": 727
+    },
+    {
+      "epoch": 0.9578947368421052,
+      "grad_norm": 0.08166198075045744,
+      "learning_rate": 0.0001721039265116285,
+      "loss": 0.2818,
+      "step": 728
+    },
+    {
+      "epoch": 0.9592105263157895,
+      "grad_norm": 0.08220017081684095,
+      "learning_rate": 0.00017199776060033997,
+      "loss": 0.2761,
+      "step": 729
+    },
+    {
+      "epoch": 0.9605263157894737,
+      "grad_norm": 0.08134420494867958,
+      "learning_rate": 0.00017189142593121993,
+      "loss": 0.2738,
+      "step": 730
+    },
+    {
+      "epoch": 0.9618421052631579,
+      "grad_norm": 0.08157316013668467,
+      "learning_rate": 0.00017178492275350958,
+      "loss": 0.2637,
+      "step": 731
+    },
+    {
+      "epoch": 0.9631578947368421,
+      "grad_norm": 0.0850556774542164,
+      "learning_rate": 0.00017167825131684513,
+      "loss": 0.292,
+      "step": 732
+    },
+    {
+      "epoch": 0.9644736842105263,
+      "grad_norm": 0.08012953024736491,
+      "learning_rate": 0.00017157141187125713,
+      "loss": 0.2723,
+      "step": 733
+    },
+    {
+      "epoch": 0.9657894736842105,
+      "grad_norm": 0.07862483655819683,
+      "learning_rate": 0.00017146440466716991,
+      "loss": 0.2856,
+      "step": 734
+    },
+    {
+      "epoch": 0.9671052631578947,
+      "grad_norm": 0.08165975689746587,
+      "learning_rate": 0.00017135722995540107,
+      "loss": 0.2729,
+      "step": 735
+    },
+    {
+      "epoch": 0.968421052631579,
+      "grad_norm": 0.07988073177015712,
+      "learning_rate": 0.00017124988798716083,
+      "loss": 0.2822,
+      "step": 736
+    },
+    {
+      "epoch": 0.9697368421052631,
+      "grad_norm": 0.08138761359544663,
+      "learning_rate": 0.00017114237901405134,
+      "loss": 0.2642,
+      "step": 737
+    },
+    {
+      "epoch": 0.9710526315789474,
+      "grad_norm": 0.08079811328305973,
+      "learning_rate": 0.0001710347032880664,
+      "loss": 0.2843,
+      "step": 738
+    },
+    {
+      "epoch": 0.9723684210526315,
+      "grad_norm": 0.08408921154912564,
+      "learning_rate": 0.00017092686106159053,
+      "loss": 0.2924,
+      "step": 739
+    },
+    {
+      "epoch": 0.9736842105263158,
+      "grad_norm": 0.0805385190287914,
+      "learning_rate": 0.00017081885258739846,
+      "loss": 0.2823,
+      "step": 740
+    },
+    {
+      "epoch": 0.975,
+      "grad_norm": 0.08475021934413124,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.2923,
+      "step": 741
+    },
+    {
+      "epoch": 0.9763157894736842,
+      "grad_norm": 0.07950603456214735,
+      "learning_rate": 0.00017060233790891296,
+      "loss": 0.275,
+      "step": 742
+    },
+    {
+      "epoch": 0.9776315789473684,
+      "grad_norm": 0.07785517335290744,
+      "learning_rate": 0.0001704938322121151,
+      "loss": 0.2831,
+      "step": 743
+    },
+    {
+      "epoch": 0.9789473684210527,
+      "grad_norm": 0.07960926560941854,
+      "learning_rate": 0.00017038516128259115,
+      "loss": 0.2871,
+      "step": 744
+    },
+    {
+      "epoch": 0.9802631578947368,
+      "grad_norm": 0.08293841281790956,
+      "learning_rate": 0.00017027632537505832,
+      "loss": 0.2753,
+      "step": 745
+    },
+    {
+      "epoch": 0.9815789473684211,
+      "grad_norm": 0.07747002473316371,
+      "learning_rate": 0.00017016732474462056,
+      "loss": 0.2839,
+      "step": 746
+    },
+    {
+      "epoch": 0.9828947368421053,
+      "grad_norm": 0.08046473275537684,
+      "learning_rate": 0.00017005815964676787,
+      "loss": 0.2768,
+      "step": 747
+    },
+    {
+      "epoch": 0.9842105263157894,
+      "grad_norm": 0.08364061752605101,
+      "learning_rate": 0.00016994883033737582,
+      "loss": 0.2931,
+      "step": 748
+    },
+    {
+      "epoch": 0.9855263157894737,
+      "grad_norm": 0.08050775331309325,
+      "learning_rate": 0.0001698393370727048,
+      "loss": 0.2893,
+      "step": 749
+    },
+    {
+      "epoch": 0.9868421052631579,
+      "grad_norm": 0.08002147081513335,
+      "learning_rate": 0.00016972968010939954,
+      "loss": 0.2848,
+      "step": 750
+    },
+    {
+      "epoch": 0.9881578947368421,
+      "grad_norm": 0.07808855249540297,
+      "learning_rate": 0.0001696198597044885,
+      "loss": 0.2846,
+      "step": 751
+    },
+    {
+      "epoch": 0.9894736842105263,
+      "grad_norm": 0.07784201379245488,
+      "learning_rate": 0.00016950987611538324,
+      "loss": 0.2838,
+      "step": 752
+    },
+    {
+      "epoch": 0.9907894736842106,
+      "grad_norm": 0.07950126160436796,
+      "learning_rate": 0.0001693997295998777,
+      "loss": 0.2933,
+      "step": 753
+    },
+    {
+      "epoch": 0.9921052631578947,
+      "grad_norm": 0.08013885766603035,
+      "learning_rate": 0.0001692894204161478,
+      "loss": 0.2703,
+      "step": 754
+    },
+    {
+      "epoch": 0.993421052631579,
+      "grad_norm": 0.08277409481218478,
+      "learning_rate": 0.00016917894882275075,
+      "loss": 0.2824,
+      "step": 755
+    },
+    {
+      "epoch": 0.9947368421052631,
+      "grad_norm": 0.0816211725278549,
+      "learning_rate": 0.00016906831507862443,
+      "loss": 0.2781,
+      "step": 756
+    },
+    {
+      "epoch": 0.9960526315789474,
+      "grad_norm": 0.08101924988182123,
+      "learning_rate": 0.00016895751944308679,
+      "loss": 0.284,
+      "step": 757
+    },
+    {
+      "epoch": 0.9973684210526316,
+      "grad_norm": 0.08179919265807176,
+      "learning_rate": 0.00016884656217583518,
+      "loss": 0.2847,
+      "step": 758
+    },
+    {
+      "epoch": 0.9986842105263158,
+      "grad_norm": 0.07951765088912952,
+      "learning_rate": 0.00016873544353694588,
+      "loss": 0.2869,
+      "step": 759
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.07809652892973841,
+      "learning_rate": 0.0001686241637868734,
+      "loss": 0.2809,
+      "step": 760
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.2806478440761566,
+      "eval_runtime": 142.699,
+      "eval_samples_per_second": 35.866,
+      "eval_steps_per_second": 1.121,
+      "step": 760
+    },
+    {
+      "epoch": 1.0013157894736842,
+      "grad_norm": 0.08089780903924765,
+      "learning_rate": 0.0001685127231864498,
+      "loss": 0.273,
+      "step": 761
+    },
+    {
+      "epoch": 1.0026315789473683,
+      "grad_norm": 0.0791344049462135,
+      "learning_rate": 0.00016840112199688432,
+      "loss": 0.2746,
+      "step": 762
+    },
+    {
+      "epoch": 1.0039473684210527,
+      "grad_norm": 0.07798719813102083,
+      "learning_rate": 0.00016828936047976248,
+      "loss": 0.2745,
+      "step": 763
+    },
+    {
+      "epoch": 1.0052631578947369,
+      "grad_norm": 0.08227033417121575,
+      "learning_rate": 0.00016817743889704565,
+      "loss": 0.2617,
+      "step": 764
+    },
+    {
+      "epoch": 1.006578947368421,
+      "grad_norm": 0.08639411451972336,
+      "learning_rate": 0.00016806535751107037,
+      "loss": 0.2674,
+      "step": 765
+    },
+    {
+      "epoch": 1.0078947368421052,
+      "grad_norm": 0.08367139936296988,
+      "learning_rate": 0.00016795311658454777,
+      "loss": 0.279,
+      "step": 766
+    },
+    {
+      "epoch": 1.0092105263157896,
+      "grad_norm": 0.08766002313298543,
+      "learning_rate": 0.00016784071638056285,
+      "loss": 0.283,
+      "step": 767
+    },
+    {
+      "epoch": 1.0105263157894737,
+      "grad_norm": 0.08427710263543763,
+      "learning_rate": 0.00016772815716257412,
+      "loss": 0.2663,
+      "step": 768
+    },
+    {
+      "epoch": 1.0118421052631579,
+      "grad_norm": 0.07801065155963043,
+      "learning_rate": 0.0001676154391944126,
+      "loss": 0.2559,
+      "step": 769
+    },
+    {
+      "epoch": 1.013157894736842,
+      "grad_norm": 0.07698562715479776,
+      "learning_rate": 0.00016750256274028152,
+      "loss": 0.2748,
+      "step": 770
+    },
+    {
+      "epoch": 1.0144736842105264,
+      "grad_norm": 0.08054717072786537,
+      "learning_rate": 0.0001673895280647556,
+      "loss": 0.2719,
+      "step": 771
+    },
+    {
+      "epoch": 1.0157894736842106,
+      "grad_norm": 0.08399901308359886,
+      "learning_rate": 0.0001672763354327804,
+      "loss": 0.2735,
+      "step": 772
+    },
+    {
+      "epoch": 1.0171052631578947,
+      "grad_norm": 0.07991338180666516,
+      "learning_rate": 0.0001671629851096717,
+      "loss": 0.2688,
+      "step": 773
+    },
+    {
+      "epoch": 1.018421052631579,
+      "grad_norm": 0.0817891282862518,
+      "learning_rate": 0.00016704947736111492,
+      "loss": 0.2695,
+      "step": 774
+    },
+    {
+      "epoch": 1.019736842105263,
+      "grad_norm": 0.084530202074618,
+      "learning_rate": 0.00016693581245316442,
+      "loss": 0.2667,
+      "step": 775
+    },
+    {
+      "epoch": 1.0210526315789474,
+      "grad_norm": 0.08294234750514276,
+      "learning_rate": 0.00016682199065224307,
+      "loss": 0.2728,
+      "step": 776
+    },
+    {
+      "epoch": 1.0223684210526316,
+      "grad_norm": 0.0832171550305921,
+      "learning_rate": 0.00016670801222514134,
+      "loss": 0.2778,
+      "step": 777
+    },
+    {
+      "epoch": 1.0236842105263158,
+      "grad_norm": 0.08255803178852009,
+      "learning_rate": 0.00016659387743901685,
+      "loss": 0.2617,
+      "step": 778
+    },
+    {
+      "epoch": 1.025,
+      "grad_norm": 0.08205321961071378,
+      "learning_rate": 0.00016647958656139378,
+      "loss": 0.2717,
+      "step": 779
+    },
+    {
+      "epoch": 1.0263157894736843,
+      "grad_norm": 0.08258890569995453,
+      "learning_rate": 0.00016636513986016213,
+      "loss": 0.2658,
+      "step": 780
+    },
+    {
+      "epoch": 1.0276315789473685,
+      "grad_norm": 0.08071843114802557,
+      "learning_rate": 0.0001662505376035772,
+      "loss": 0.2613,
+      "step": 781
+    },
+    {
+      "epoch": 1.0289473684210526,
+      "grad_norm": 0.07886987321903878,
+      "learning_rate": 0.00016613578006025872,
+      "loss": 0.257,
+      "step": 782
+    },
+    {
+      "epoch": 1.0302631578947368,
+      "grad_norm": 0.07877983241946228,
+      "learning_rate": 0.00016602086749919063,
+      "loss": 0.269,
+      "step": 783
+    },
+    {
+      "epoch": 1.0315789473684212,
+      "grad_norm": 0.08280631625375447,
+      "learning_rate": 0.0001659058001897201,
+      "loss": 0.2687,
+      "step": 784
+    },
+    {
+      "epoch": 1.0328947368421053,
+      "grad_norm": 0.08033744101706376,
+      "learning_rate": 0.00016579057840155703,
+      "loss": 0.2676,
+      "step": 785
+    },
+    {
+      "epoch": 1.0342105263157895,
+      "grad_norm": 0.07853625991786962,
+      "learning_rate": 0.00016567520240477344,
+      "loss": 0.2475,
+      "step": 786
+    },
+    {
+      "epoch": 1.0355263157894736,
+      "grad_norm": 0.08321225528586546,
+      "learning_rate": 0.00016555967246980276,
+      "loss": 0.2733,
+      "step": 787
+    },
+    {
+      "epoch": 1.0368421052631578,
+      "grad_norm": 0.08271091060501498,
+      "learning_rate": 0.00016544398886743933,
+      "loss": 0.2669,
+      "step": 788
+    },
+    {
+      "epoch": 1.0381578947368422,
+      "grad_norm": 0.08480813500178941,
+      "learning_rate": 0.00016532815186883748,
+      "loss": 0.2702,
+      "step": 789
+    },
+    {
+      "epoch": 1.0394736842105263,
+      "grad_norm": 0.08423735368240373,
+      "learning_rate": 0.0001652121617455113,
+      "loss": 0.266,
+      "step": 790
+    },
+    {
+      "epoch": 1.0407894736842105,
+      "grad_norm": 0.08270665791388272,
+      "learning_rate": 0.00016509601876933374,
+      "loss": 0.2748,
+      "step": 791
+    },
+    {
+      "epoch": 1.0421052631578946,
+      "grad_norm": 0.08074688262914587,
+      "learning_rate": 0.000164979723212536,
+      "loss": 0.2681,
+      "step": 792
+    },
+    {
+      "epoch": 1.043421052631579,
+      "grad_norm": 0.08329297167987026,
+      "learning_rate": 0.0001648632753477068,
+      "loss": 0.2737,
+      "step": 793
+    },
+    {
+      "epoch": 1.0447368421052632,
+      "grad_norm": 0.08195503131845851,
+      "learning_rate": 0.0001647466754477921,
+      "loss": 0.2605,
+      "step": 794
+    },
+    {
+      "epoch": 1.0460526315789473,
+      "grad_norm": 0.08219664275017928,
+      "learning_rate": 0.00016462992378609407,
+      "loss": 0.2726,
+      "step": 795
+    },
+    {
+      "epoch": 1.0473684210526315,
+      "grad_norm": 0.081155677466978,
+      "learning_rate": 0.00016451302063627066,
+      "loss": 0.2683,
+      "step": 796
+    },
+    {
+      "epoch": 1.0486842105263159,
+      "grad_norm": 0.07978496536582978,
+      "learning_rate": 0.0001643959662723348,
+      "loss": 0.2639,
+      "step": 797
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.07736573715270778,
+      "learning_rate": 0.00016427876096865394,
+      "loss": 0.2523,
+      "step": 798
+    },
+    {
+      "epoch": 1.0513157894736842,
+      "grad_norm": 0.08461912280741266,
+      "learning_rate": 0.0001641614049999493,
+      "loss": 0.269,
+      "step": 799
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.08431583636176626,
+      "learning_rate": 0.00016404389864129533,
+      "loss": 0.2563,
+      "step": 800
+    },
+    {
+      "epoch": 1.0539473684210527,
+      "grad_norm": 0.08694565310464801,
+      "learning_rate": 0.00016392624216811879,
+      "loss": 0.2778,
+      "step": 801
+    },
+    {
+      "epoch": 1.055263157894737,
+      "grad_norm": 0.08352264819913631,
+      "learning_rate": 0.00016380843585619845,
+      "loss": 0.2742,
+      "step": 802
+    },
+    {
+      "epoch": 1.056578947368421,
+      "grad_norm": 0.08421963392115288,
+      "learning_rate": 0.0001636904799816643,
+      "loss": 0.271,
+      "step": 803
+    },
+    {
+      "epoch": 1.0578947368421052,
+      "grad_norm": 0.07881294236501658,
+      "learning_rate": 0.00016357237482099684,
+      "loss": 0.2599,
+      "step": 804
+    },
+    {
+      "epoch": 1.0592105263157894,
+      "grad_norm": 0.08602162394576658,
+      "learning_rate": 0.0001634541206510264,
+      "loss": 0.2729,
+      "step": 805
+    },
+    {
+      "epoch": 1.0605263157894738,
+      "grad_norm": 0.08348451901525375,
+      "learning_rate": 0.00016333571774893285,
+      "loss": 0.2753,
+      "step": 806
+    },
+    {
+      "epoch": 1.061842105263158,
+      "grad_norm": 0.0818737526419465,
+      "learning_rate": 0.00016321716639224434,
+      "loss": 0.2715,
+      "step": 807
+    },
+    {
+      "epoch": 1.063157894736842,
+      "grad_norm": 0.08286281227398601,
+      "learning_rate": 0.00016309846685883726,
+      "loss": 0.2675,
+      "step": 808
+    },
+    {
+      "epoch": 1.0644736842105262,
+      "grad_norm": 0.07966839196252047,
+      "learning_rate": 0.00016297961942693512,
+      "loss": 0.2558,
+      "step": 809
+    },
+    {
+      "epoch": 1.0657894736842106,
+      "grad_norm": 0.08414413297914877,
+      "learning_rate": 0.0001628606243751082,
+      "loss": 0.2844,
+      "step": 810
+    },
+    {
+      "epoch": 1.0671052631578948,
+      "grad_norm": 0.08555085876337333,
+      "learning_rate": 0.00016274148198227282,
+      "loss": 0.2851,
+      "step": 811
+    },
+    {
+      "epoch": 1.068421052631579,
+      "grad_norm": 0.08082783425810476,
+      "learning_rate": 0.00016262219252769064,
+      "loss": 0.2646,
+      "step": 812
+    },
+    {
+      "epoch": 1.069736842105263,
+      "grad_norm": 0.08197732764454009,
+      "learning_rate": 0.00016250275629096786,
+      "loss": 0.2779,
+      "step": 813
+    },
+    {
+      "epoch": 1.0710526315789473,
+      "grad_norm": 0.08526782512435523,
+      "learning_rate": 0.00016238317355205494,
+      "loss": 0.2799,
+      "step": 814
+    },
+    {
+      "epoch": 1.0723684210526316,
+      "grad_norm": 0.07823583584470825,
+      "learning_rate": 0.00016226344459124566,
+      "loss": 0.2643,
+      "step": 815
+    },
+    {
+      "epoch": 1.0736842105263158,
+      "grad_norm": 0.08603151669795409,
+      "learning_rate": 0.00016214356968917648,
+      "loss": 0.273,
+      "step": 816
+    },
+    {
+      "epoch": 1.075,
+      "grad_norm": 0.08851720457464476,
+      "learning_rate": 0.000162023549126826,
+      "loss": 0.2721,
+      "step": 817
+    },
+    {
+      "epoch": 1.0763157894736841,
+      "grad_norm": 0.08133793366760693,
+      "learning_rate": 0.00016190338318551427,
+      "loss": 0.255,
+      "step": 818
+    },
+    {
+      "epoch": 1.0776315789473685,
+      "grad_norm": 0.082819100026463,
+      "learning_rate": 0.00016178307214690193,
+      "loss": 0.2624,
+      "step": 819
+    },
+    {
+      "epoch": 1.0789473684210527,
+      "grad_norm": 0.08530141351749654,
+      "learning_rate": 0.00016166261629298995,
+      "loss": 0.2659,
+      "step": 820
+    },
+    {
+      "epoch": 1.0802631578947368,
+      "grad_norm": 0.08366548952005705,
+      "learning_rate": 0.00016154201590611852,
+      "loss": 0.2627,
+      "step": 821
+    },
+    {
+      "epoch": 1.081578947368421,
+      "grad_norm": 0.08224916068858729,
+      "learning_rate": 0.0001614212712689668,
+      "loss": 0.2702,
+      "step": 822
+    },
+    {
+      "epoch": 1.0828947368421054,
+      "grad_norm": 0.08321699362162292,
+      "learning_rate": 0.0001613003826645519,
+      "loss": 0.2668,
+      "step": 823
+    },
+    {
+      "epoch": 1.0842105263157895,
+      "grad_norm": 0.07940176654017943,
+      "learning_rate": 0.0001611793503762285,
+      "loss": 0.2666,
+      "step": 824
+    },
+    {
+      "epoch": 1.0855263157894737,
+      "grad_norm": 0.0810039782214103,
+      "learning_rate": 0.00016105817468768798,
+      "loss": 0.2748,
+      "step": 825
+    },
+    {
+      "epoch": 1.0868421052631578,
+      "grad_norm": 0.08080799943317177,
+      "learning_rate": 0.00016093685588295786,
+      "loss": 0.2664,
+      "step": 826
+    },
+    {
+      "epoch": 1.0881578947368422,
+      "grad_norm": 0.07963385945239661,
+      "learning_rate": 0.00016081539424640118,
+      "loss": 0.279,
+      "step": 827
+    },
+    {
+      "epoch": 1.0894736842105264,
+      "grad_norm": 0.08180594063442842,
+      "learning_rate": 0.00016069379006271566,
+      "loss": 0.2735,
+      "step": 828
+    },
+    {
+      "epoch": 1.0907894736842105,
+      "grad_norm": 0.08134763460779691,
+      "learning_rate": 0.00016057204361693327,
+      "loss": 0.2666,
+      "step": 829
+    },
+    {
+      "epoch": 1.0921052631578947,
+      "grad_norm": 0.08433241545175918,
+      "learning_rate": 0.0001604501551944193,
+      "loss": 0.2582,
+      "step": 830
+    },
+    {
+      "epoch": 1.0934210526315788,
+      "grad_norm": 0.08109132578451536,
+      "learning_rate": 0.0001603281250808719,
+      "loss": 0.2648,
+      "step": 831
+    },
+    {
+      "epoch": 1.0947368421052632,
+      "grad_norm": 0.09007590757575104,
+      "learning_rate": 0.00016020595356232135,
+      "loss": 0.2681,
+      "step": 832
+    },
+    {
+      "epoch": 1.0960526315789474,
+      "grad_norm": 0.08677922191238775,
+      "learning_rate": 0.00016008364092512926,
+      "loss": 0.2761,
+      "step": 833
+    },
+    {
+      "epoch": 1.0973684210526315,
+      "grad_norm": 0.08575921288664821,
+      "learning_rate": 0.00015996118745598817,
+      "loss": 0.2696,
+      "step": 834
+    },
+    {
+      "epoch": 1.0986842105263157,
+      "grad_norm": 0.08041317041434254,
+      "learning_rate": 0.00015983859344192061,
+      "loss": 0.2689,
+      "step": 835
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.08065596311704078,
+      "learning_rate": 0.00015971585917027862,
+      "loss": 0.2687,
+      "step": 836
+    },
+    {
+      "epoch": 1.1013157894736842,
+      "grad_norm": 0.0790113706176156,
+      "learning_rate": 0.00015959298492874288,
+      "loss": 0.2555,
+      "step": 837
+    },
+    {
+      "epoch": 1.1026315789473684,
+      "grad_norm": 0.08012262373177113,
+      "learning_rate": 0.0001594699710053223,
+      "loss": 0.2685,
+      "step": 838
+    },
+    {
+      "epoch": 1.1039473684210526,
+      "grad_norm": 0.08504138731346632,
+      "learning_rate": 0.00015934681768835297,
+      "loss": 0.2805,
+      "step": 839
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.08095254397891562,
+      "learning_rate": 0.00015922352526649803,
+      "loss": 0.2709,
+      "step": 840
+    },
+    {
+      "epoch": 1.106578947368421,
+      "grad_norm": 0.08674994051804193,
+      "learning_rate": 0.00015910009402874631,
+      "loss": 0.2806,
+      "step": 841
+    },
+    {
+      "epoch": 1.1078947368421053,
+      "grad_norm": 0.09194758435288344,
+      "learning_rate": 0.0001589765242644124,
+      "loss": 0.2736,
+      "step": 842
+    },
+    {
+      "epoch": 1.1092105263157894,
+      "grad_norm": 0.08318187378912749,
+      "learning_rate": 0.00015885281626313517,
+      "loss": 0.2606,
+      "step": 843
+    },
+    {
+      "epoch": 1.1105263157894736,
+      "grad_norm": 0.0825846308272865,
+      "learning_rate": 0.00015872897031487791,
+      "loss": 0.2628,
+      "step": 844
+    },
+    {
+      "epoch": 1.111842105263158,
+      "grad_norm": 0.08112434803943486,
+      "learning_rate": 0.00015860498670992691,
+      "loss": 0.2793,
+      "step": 845
+    },
+    {
+      "epoch": 1.1131578947368421,
+      "grad_norm": 0.08537425779356113,
+      "learning_rate": 0.00015848086573889137,
+      "loss": 0.2779,
+      "step": 846
+    },
+    {
+      "epoch": 1.1144736842105263,
+      "grad_norm": 0.08315694519509176,
+      "learning_rate": 0.00015835660769270232,
+      "loss": 0.2719,
+      "step": 847
+    },
+    {
+      "epoch": 1.1157894736842104,
+      "grad_norm": 0.08130780896158304,
+      "learning_rate": 0.00015823221286261215,
+      "loss": 0.2653,
+      "step": 848
+    },
+    {
+      "epoch": 1.1171052631578948,
+      "grad_norm": 0.08219491647444532,
+      "learning_rate": 0.00015810768154019385,
+      "loss": 0.2765,
+      "step": 849
+    },
+    {
+      "epoch": 1.118421052631579,
+      "grad_norm": 0.08240059160474296,
+      "learning_rate": 0.0001579830140173403,
+      "loss": 0.2737,
+      "step": 850
+    },
+    {
+      "epoch": 1.1197368421052631,
+      "grad_norm": 0.083577356578667,
+      "learning_rate": 0.00015785821058626366,
+      "loss": 0.2661,
+      "step": 851
+    },
+    {
+      "epoch": 1.1210526315789473,
+      "grad_norm": 0.07834322980894085,
+      "learning_rate": 0.00015773327153949465,
+      "loss": 0.2729,
+      "step": 852
+    },
+    {
+      "epoch": 1.1223684210526317,
+      "grad_norm": 0.08346738623736531,
+      "learning_rate": 0.00015760819716988187,
+      "loss": 0.275,
+      "step": 853
+    },
+    {
+      "epoch": 1.1236842105263158,
+      "grad_norm": 0.07884897152428384,
+      "learning_rate": 0.00015748298777059112,
+      "loss": 0.2805,
+      "step": 854
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 0.08020439719046484,
+      "learning_rate": 0.0001573576436351046,
+      "loss": 0.2587,
+      "step": 855
+    },
+    {
+      "epoch": 1.1263157894736842,
+      "grad_norm": 0.08327214298097156,
+      "learning_rate": 0.0001572321650572205,
+      "loss": 0.2626,
+      "step": 856
+    },
+    {
+      "epoch": 1.1276315789473683,
+      "grad_norm": 0.08372000718554913,
+      "learning_rate": 0.00015710655233105194,
+      "loss": 0.2743,
+      "step": 857
+    },
+    {
+      "epoch": 1.1289473684210527,
+      "grad_norm": 0.08197882426728544,
+      "learning_rate": 0.00015698080575102661,
+      "loss": 0.2657,
+      "step": 858
+    },
+    {
+      "epoch": 1.1302631578947369,
+      "grad_norm": 0.08271785330561393,
+      "learning_rate": 0.00015685492561188594,
+      "loss": 0.2696,
+      "step": 859
+    },
+    {
+      "epoch": 1.131578947368421,
+      "grad_norm": 0.08153102919965115,
+      "learning_rate": 0.00015672891220868432,
+      "loss": 0.2627,
+      "step": 860
+    },
+    {
+      "epoch": 1.1328947368421052,
+      "grad_norm": 0.08273222452553464,
+      "learning_rate": 0.00015660276583678853,
+      "loss": 0.2742,
+      "step": 861
+    },
+    {
+      "epoch": 1.1342105263157896,
+      "grad_norm": 0.08149761866108174,
+      "learning_rate": 0.0001564764867918771,
+      "loss": 0.284,
+      "step": 862
+    },
+    {
+      "epoch": 1.1355263157894737,
+      "grad_norm": 0.07732037718328716,
+      "learning_rate": 0.0001563500753699395,
+      "loss": 0.2559,
+      "step": 863
+    },
+    {
+      "epoch": 1.1368421052631579,
+      "grad_norm": 0.08033074721619064,
+      "learning_rate": 0.00015622353186727544,
+      "loss": 0.2651,
+      "step": 864
+    },
+    {
+      "epoch": 1.138157894736842,
+      "grad_norm": 0.07988808525323399,
+      "learning_rate": 0.0001560968565804942,
+      "loss": 0.2676,
+      "step": 865
+    },
+    {
+      "epoch": 1.1394736842105262,
+      "grad_norm": 0.07997455834407387,
+      "learning_rate": 0.00015597004980651407,
+      "loss": 0.2597,
+      "step": 866
+    },
+    {
+      "epoch": 1.1407894736842106,
+      "grad_norm": 0.07985601487062838,
+      "learning_rate": 0.0001558431118425614,
+      "loss": 0.257,
+      "step": 867
+    },
+    {
+      "epoch": 1.1421052631578947,
+      "grad_norm": 0.08269154375775314,
+      "learning_rate": 0.0001557160429861702,
+      "loss": 0.2807,
+      "step": 868
+    },
+    {
+      "epoch": 1.143421052631579,
+      "grad_norm": 0.08357702138558408,
+      "learning_rate": 0.00015558884353518107,
+      "loss": 0.2767,
+      "step": 869
+    },
+    {
+      "epoch": 1.1447368421052633,
+      "grad_norm": 0.08414408194028052,
+      "learning_rate": 0.00015546151378774086,
+      "loss": 0.274,
+      "step": 870
+    },
+    {
+      "epoch": 1.1460526315789474,
+      "grad_norm": 0.08701268358835322,
+      "learning_rate": 0.00015533405404230188,
+      "loss": 0.2811,
+      "step": 871
+    },
+    {
+      "epoch": 1.1473684210526316,
+      "grad_norm": 0.08262282790058861,
+      "learning_rate": 0.000155206464597621,
+      "loss": 0.2731,
+      "step": 872
+    },
+    {
+      "epoch": 1.1486842105263158,
+      "grad_norm": 0.08408606653507734,
+      "learning_rate": 0.00015507874575275917,
+      "loss": 0.251,
+      "step": 873
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.08049323373059812,
+      "learning_rate": 0.0001549508978070806,
+      "loss": 0.2715,
+      "step": 874
+    },
+    {
+      "epoch": 1.1513157894736843,
+      "grad_norm": 0.07972519186854046,
+      "learning_rate": 0.0001548229210602522,
+      "loss": 0.2671,
+      "step": 875
+    },
+    {
+      "epoch": 1.1526315789473685,
+      "grad_norm": 0.07882804213132247,
+      "learning_rate": 0.00015469481581224272,
+      "loss": 0.2619,
+      "step": 876
+    },
+    {
+      "epoch": 1.1539473684210526,
+      "grad_norm": 0.0798538301353752,
+      "learning_rate": 0.00015456658236332203,
+      "loss": 0.2736,
+      "step": 877
+    },
+    {
+      "epoch": 1.1552631578947368,
+      "grad_norm": 0.08315715083917044,
+      "learning_rate": 0.00015443822101406064,
+      "loss": 0.2664,
+      "step": 878
+    },
+    {
+      "epoch": 1.1565789473684212,
+      "grad_norm": 0.08372404146892015,
+      "learning_rate": 0.00015430973206532878,
+      "loss": 0.2686,
+      "step": 879
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.08175706218650852,
+      "learning_rate": 0.00015418111581829574,
+      "loss": 0.2622,
+      "step": 880
+    },
+    {
+      "epoch": 1.1592105263157895,
+      "grad_norm": 0.07882686056716785,
+      "learning_rate": 0.00015405237257442924,
+      "loss": 0.269,
+      "step": 881
+    },
+    {
+      "epoch": 1.1605263157894736,
+      "grad_norm": 0.07950928567287992,
+      "learning_rate": 0.0001539235026354946,
+      "loss": 0.2628,
+      "step": 882
+    },
+    {
+      "epoch": 1.1618421052631578,
+      "grad_norm": 0.0850340095145124,
+      "learning_rate": 0.00015379450630355424,
+      "loss": 0.2791,
+      "step": 883
+    },
+    {
+      "epoch": 1.1631578947368422,
+      "grad_norm": 0.0941613058127184,
+      "learning_rate": 0.0001536653838809667,
+      "loss": 0.279,
+      "step": 884
+    },
+    {
+      "epoch": 1.1644736842105263,
+      "grad_norm": 0.08251843104601717,
+      "learning_rate": 0.00015353613567038607,
+      "loss": 0.2859,
+      "step": 885
+    },
+    {
+      "epoch": 1.1657894736842105,
+      "grad_norm": 0.08038962867293699,
+      "learning_rate": 0.0001534067619747614,
+      "loss": 0.2747,
+      "step": 886
+    },
+    {
+      "epoch": 1.1671052631578946,
+      "grad_norm": 0.0794561388070862,
+      "learning_rate": 0.00015327726309733572,
+      "loss": 0.2686,
+      "step": 887
+    },
+    {
+      "epoch": 1.168421052631579,
+      "grad_norm": 0.07884245258276419,
+      "learning_rate": 0.0001531476393416456,
+      "loss": 0.2672,
+      "step": 888
+    },
+    {
+      "epoch": 1.1697368421052632,
+      "grad_norm": 0.08136001457411983,
+      "learning_rate": 0.00015301789101152026,
+      "loss": 0.2672,
+      "step": 889
+    },
+    {
+      "epoch": 1.1710526315789473,
+      "grad_norm": 0.08416373900750189,
+      "learning_rate": 0.00015288801841108093,
+      "loss": 0.2823,
+      "step": 890
+    },
+    {
+      "epoch": 1.1723684210526315,
+      "grad_norm": 0.0820367167436313,
+      "learning_rate": 0.0001527580218447401,
+      "loss": 0.2578,
+      "step": 891
+    },
+    {
+      "epoch": 1.1736842105263159,
+      "grad_norm": 0.08309627633941746,
+      "learning_rate": 0.0001526279016172008,
+      "loss": 0.267,
+      "step": 892
+    },
+    {
+      "epoch": 1.175,
+      "grad_norm": 0.08128917587224013,
+      "learning_rate": 0.000152497658033456,
+      "loss": 0.2561,
+      "step": 893
+    },
+    {
+      "epoch": 1.1763157894736842,
+      "grad_norm": 0.08178638667577875,
+      "learning_rate": 0.00015236729139878782,
+      "loss": 0.2768,
+      "step": 894
+    },
+    {
+      "epoch": 1.1776315789473684,
+      "grad_norm": 0.0801661892105483,
+      "learning_rate": 0.0001522368020187666,
+      "loss": 0.2543,
+      "step": 895
+    },
+    {
+      "epoch": 1.1789473684210527,
+      "grad_norm": 0.08213132733512518,
+      "learning_rate": 0.00015210619019925066,
+      "loss": 0.2636,
+      "step": 896
+    },
+    {
+      "epoch": 1.180263157894737,
+      "grad_norm": 0.08080981772989361,
+      "learning_rate": 0.00015197545624638504,
+      "loss": 0.267,
+      "step": 897
+    },
+    {
+      "epoch": 1.181578947368421,
+      "grad_norm": 0.07879728180506462,
+      "learning_rate": 0.00015184460046660137,
+      "loss": 0.2622,
+      "step": 898
+    },
+    {
+      "epoch": 1.1828947368421052,
+      "grad_norm": 0.07930675022154604,
+      "learning_rate": 0.00015171362316661652,
+      "loss": 0.2635,
+      "step": 899
+    },
+    {
+      "epoch": 1.1842105263157894,
+      "grad_norm": 0.0827303133693508,
+      "learning_rate": 0.00015158252465343242,
+      "loss": 0.2731,
+      "step": 900
+    },
+    {
+      "epoch": 1.1855263157894738,
+      "grad_norm": 0.08133340547715927,
+      "learning_rate": 0.00015145130523433492,
+      "loss": 0.2691,
+      "step": 901
+    },
+    {
+      "epoch": 1.186842105263158,
+      "grad_norm": 0.08320419791917552,
+      "learning_rate": 0.00015131996521689352,
+      "loss": 0.2831,
+      "step": 902
+    },
+    {
+      "epoch": 1.188157894736842,
+      "grad_norm": 0.078583719040117,
+      "learning_rate": 0.00015118850490896012,
+      "loss": 0.2605,
+      "step": 903
+    },
+    {
+      "epoch": 1.1894736842105262,
+      "grad_norm": 0.07987941146339982,
+      "learning_rate": 0.00015105692461866874,
+      "loss": 0.271,
+      "step": 904
+    },
+    {
+      "epoch": 1.1907894736842106,
+      "grad_norm": 0.08065385126525215,
+      "learning_rate": 0.0001509252246544346,
+      "loss": 0.2748,
+      "step": 905
+    },
+    {
+      "epoch": 1.1921052631578948,
+      "grad_norm": 0.08045853100208103,
+      "learning_rate": 0.00015079340532495343,
+      "loss": 0.2793,
+      "step": 906
+    },
+    {
+      "epoch": 1.193421052631579,
+      "grad_norm": 0.07925415373688983,
+      "learning_rate": 0.00015066146693920072,
+      "loss": 0.2625,
+      "step": 907
+    },
+    {
+      "epoch": 1.194736842105263,
+      "grad_norm": 0.08327329284089614,
+      "learning_rate": 0.000150529409806431,
+      "loss": 0.2679,
+      "step": 908
+    },
+    {
+      "epoch": 1.1960526315789473,
+      "grad_norm": 0.08369037313818105,
+      "learning_rate": 0.0001503972342361772,
+      "loss": 0.2729,
+      "step": 909
+    },
+    {
+      "epoch": 1.1973684210526316,
+      "grad_norm": 0.08698590728820178,
+      "learning_rate": 0.00015026494053824982,
+      "loss": 0.2792,
+      "step": 910
+    },
+    {
+      "epoch": 1.1986842105263158,
+      "grad_norm": 0.08234237296181915,
+      "learning_rate": 0.0001501325290227362,
+      "loss": 0.2723,
+      "step": 911
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.08018132065204943,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.2582,
+      "step": 912
+    },
+    {
+      "epoch": 1.2013157894736841,
+      "grad_norm": 0.08029223756152773,
+      "learning_rate": 0.0001498673537806801,
+      "loss": 0.2605,
+      "step": 913
+    },
+    {
+      "epoch": 1.2026315789473685,
+      "grad_norm": 0.07906404673525916,
+      "learning_rate": 0.00014973459067569022,
+      "loss": 0.2691,
+      "step": 914
+    },
+    {
+      "epoch": 1.2039473684210527,
+      "grad_norm": 0.0838278381246816,
+      "learning_rate": 0.00014960171099621795,
+      "loss": 0.2697,
+      "step": 915
+    },
+    {
+      "epoch": 1.2052631578947368,
+      "grad_norm": 0.08289248581844369,
+      "learning_rate": 0.00014946871505372425,
+      "loss": 0.2789,
+      "step": 916
+    },
+    {
+      "epoch": 1.206578947368421,
+      "grad_norm": 0.08065473943197897,
+      "learning_rate": 0.0001493356031599425,
+      "loss": 0.2741,
+      "step": 917
+    },
+    {
+      "epoch": 1.2078947368421054,
+      "grad_norm": 0.08222068877129116,
+      "learning_rate": 0.00014920237562687785,
+      "loss": 0.268,
+      "step": 918
+    },
+    {
+      "epoch": 1.2092105263157895,
+      "grad_norm": 0.0819456615892803,
+      "learning_rate": 0.00014906903276680654,
+      "loss": 0.267,
+      "step": 919
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.08178515229941181,
+      "learning_rate": 0.00014893557489227517,
+      "loss": 0.2711,
+      "step": 920
+    },
+    {
+      "epoch": 1.2118421052631578,
+      "grad_norm": 0.08167974669157664,
+      "learning_rate": 0.00014880200231609983,
+      "loss": 0.2623,
+      "step": 921
+    },
+    {
+      "epoch": 1.2131578947368422,
+      "grad_norm": 0.07644033108650616,
+      "learning_rate": 0.00014866831535136554,
+      "loss": 0.2592,
+      "step": 922
+    },
+    {
+      "epoch": 1.2144736842105264,
+      "grad_norm": 0.08247670698556289,
+      "learning_rate": 0.00014853451431142537,
+      "loss": 0.2563,
+      "step": 923
+    },
+    {
+      "epoch": 1.2157894736842105,
+      "grad_norm": 0.08121309159730847,
+      "learning_rate": 0.0001484005995098999,
+      "loss": 0.2651,
+      "step": 924
+    },
+    {
+      "epoch": 1.2171052631578947,
+      "grad_norm": 0.0813185941928777,
+      "learning_rate": 0.0001482665712606762,
+      "loss": 0.2654,
+      "step": 925
+    },
+    {
+      "epoch": 1.2184210526315788,
+      "grad_norm": 0.07992066628575906,
+      "learning_rate": 0.00014813242987790734,
+      "loss": 0.2601,
+      "step": 926
+    },
+    {
+      "epoch": 1.2197368421052632,
+      "grad_norm": 0.08348022762772563,
+      "learning_rate": 0.00014799817567601157,
+      "loss": 0.2698,
+      "step": 927
+    },
+    {
+      "epoch": 1.2210526315789474,
+      "grad_norm": 0.08164215203446346,
+      "learning_rate": 0.0001478638089696716,
+      "loss": 0.261,
+      "step": 928
+    },
+    {
+      "epoch": 1.2223684210526315,
+      "grad_norm": 0.08053428830419256,
+      "learning_rate": 0.00014772933007383372,
+      "loss": 0.271,
+      "step": 929
+    },
+    {
+      "epoch": 1.2236842105263157,
+      "grad_norm": 0.0855332022802314,
+      "learning_rate": 0.00014759473930370736,
+      "loss": 0.2806,
+      "step": 930
+    },
+    {
+      "epoch": 1.225,
+      "grad_norm": 0.07841552340055463,
+      "learning_rate": 0.00014746003697476404,
+      "loss": 0.2562,
+      "step": 931
+    },
+    {
+      "epoch": 1.2263157894736842,
+      "grad_norm": 0.08012970553706716,
+      "learning_rate": 0.00014732522340273684,
+      "loss": 0.2653,
+      "step": 932
+    },
+    {
+      "epoch": 1.2276315789473684,
+      "grad_norm": 0.08013137552721004,
+      "learning_rate": 0.00014719029890361955,
+      "loss": 0.2614,
+      "step": 933
+    },
+    {
+      "epoch": 1.2289473684210526,
+      "grad_norm": 0.08076902086359555,
+      "learning_rate": 0.000147055263793666,
+      "loss": 0.2726,
+      "step": 934
+    },
+    {
+      "epoch": 1.2302631578947367,
+      "grad_norm": 0.07816751188820396,
+      "learning_rate": 0.0001469201183893892,
+      "loss": 0.2697,
+      "step": 935
+    },
+    {
+      "epoch": 1.231578947368421,
+      "grad_norm": 0.07958991637658457,
+      "learning_rate": 0.0001467848630075608,
+      "loss": 0.2718,
+      "step": 936
+    },
+    {
+      "epoch": 1.2328947368421053,
+      "grad_norm": 0.07907729388333563,
+      "learning_rate": 0.00014664949796521013,
+      "loss": 0.2542,
+      "step": 937
+    },
+    {
+      "epoch": 1.2342105263157894,
+      "grad_norm": 0.08337394014823166,
+      "learning_rate": 0.00014651402357962367,
+      "loss": 0.2634,
+      "step": 938
+    },
+    {
+      "epoch": 1.2355263157894738,
+      "grad_norm": 0.07981852744305896,
+      "learning_rate": 0.00014637844016834406,
+      "loss": 0.2669,
+      "step": 939
+    },
+    {
+      "epoch": 1.236842105263158,
+      "grad_norm": 0.07805563922628504,
+      "learning_rate": 0.00014624274804916958,
+      "loss": 0.2604,
+      "step": 940
+    },
+    {
+      "epoch": 1.2381578947368421,
+      "grad_norm": 0.07875476925853715,
+      "learning_rate": 0.00014610694754015326,
+      "loss": 0.2605,
+      "step": 941
+    },
+    {
+      "epoch": 1.2394736842105263,
+      "grad_norm": 0.08212032555073835,
+      "learning_rate": 0.00014597103895960226,
+      "loss": 0.2608,
+      "step": 942
+    },
+    {
+      "epoch": 1.2407894736842104,
+      "grad_norm": 0.08193583484443283,
+      "learning_rate": 0.00014583502262607696,
+      "loss": 0.2671,
+      "step": 943
+    },
+    {
+      "epoch": 1.2421052631578948,
+      "grad_norm": 0.08018625970236232,
+      "learning_rate": 0.00014569889885839037,
+      "loss": 0.2482,
+      "step": 944
+    },
+    {
+      "epoch": 1.243421052631579,
+      "grad_norm": 0.08080487708321592,
+      "learning_rate": 0.00014556266797560732,
+      "loss": 0.2581,
+      "step": 945
+    },
+    {
+      "epoch": 1.2447368421052631,
+      "grad_norm": 0.08127791305985531,
+      "learning_rate": 0.00014542633029704366,
+      "loss": 0.2645,
+      "step": 946
+    },
+    {
+      "epoch": 1.2460526315789473,
+      "grad_norm": 0.0789744299917556,
+      "learning_rate": 0.00014528988614226563,
+      "loss": 0.258,
+      "step": 947
+    },
+    {
+      "epoch": 1.2473684210526317,
+      "grad_norm": 0.08138536304735684,
+      "learning_rate": 0.00014515333583108896,
+      "loss": 0.2602,
+      "step": 948
+    },
+    {
+      "epoch": 1.2486842105263158,
+      "grad_norm": 0.08347086931054207,
+      "learning_rate": 0.00014501667968357825,
+      "loss": 0.2737,
+      "step": 949
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.07961359280977227,
+      "learning_rate": 0.00014487991802004623,
+      "loss": 0.269,
+      "step": 950
+    },
+    {
+      "epoch": 1.2513157894736842,
+      "grad_norm": 0.08326297319920578,
+      "learning_rate": 0.00014474305116105284,
+      "loss": 0.2627,
+      "step": 951
+    },
+    {
+      "epoch": 1.2526315789473683,
+      "grad_norm": 0.08085513791177462,
+      "learning_rate": 0.00014460607942740468,
+      "loss": 0.2679,
+      "step": 952
+    },
+    {
+      "epoch": 1.2539473684210527,
+      "grad_norm": 0.07868241470750258,
+      "learning_rate": 0.00014446900314015411,
+      "loss": 0.2699,
+      "step": 953
+    },
+    {
+      "epoch": 1.2552631578947369,
+      "grad_norm": 0.0836320162619728,
+      "learning_rate": 0.0001443318226205986,
+      "loss": 0.2792,
+      "step": 954
+    },
+    {
+      "epoch": 1.256578947368421,
+      "grad_norm": 0.08451883775470095,
+      "learning_rate": 0.00014419453819027988,
+      "loss": 0.2739,
+      "step": 955
+    },
+    {
+      "epoch": 1.2578947368421054,
+      "grad_norm": 0.08034662053541705,
+      "learning_rate": 0.00014405715017098335,
+      "loss": 0.263,
+      "step": 956
+    },
+    {
+      "epoch": 1.2592105263157896,
+      "grad_norm": 0.07928448098795568,
+      "learning_rate": 0.00014391965888473703,
+      "loss": 0.2543,
+      "step": 957
+    },
+    {
+      "epoch": 1.2605263157894737,
+      "grad_norm": 0.08392429826580132,
+      "learning_rate": 0.0001437820646538112,
+      "loss": 0.2775,
+      "step": 958
+    },
+    {
+      "epoch": 1.2618421052631579,
+      "grad_norm": 0.07940539289961981,
+      "learning_rate": 0.00014364436780071727,
+      "loss": 0.2685,
+      "step": 959
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.08036645877207896,
+      "learning_rate": 0.00014350656864820733,
+      "loss": 0.2683,
+      "step": 960
+    },
+    {
+      "epoch": 1.2644736842105262,
+      "grad_norm": 0.07812397442869837,
+      "learning_rate": 0.0001433686675192731,
+      "loss": 0.2639,
+      "step": 961
+    },
+    {
+      "epoch": 1.2657894736842106,
+      "grad_norm": 0.07914775180391949,
+      "learning_rate": 0.00014323066473714542,
+      "loss": 0.2612,
+      "step": 962
+    },
+    {
+      "epoch": 1.2671052631578947,
+      "grad_norm": 0.07880507910786637,
+      "learning_rate": 0.00014309256062529344,
+      "loss": 0.269,
+      "step": 963
+    },
+    {
+      "epoch": 1.268421052631579,
+      "grad_norm": 0.08099065681511952,
+      "learning_rate": 0.0001429543555074237,
+      "loss": 0.2649,
+      "step": 964
+    },
+    {
+      "epoch": 1.2697368421052633,
+      "grad_norm": 0.07889461321025316,
+      "learning_rate": 0.00014281604970747954,
+      "loss": 0.2618,
+      "step": 965
+    },
+    {
+      "epoch": 1.2710526315789474,
+      "grad_norm": 0.07976753992813429,
+      "learning_rate": 0.00014267764354964038,
+      "loss": 0.2706,
+      "step": 966
+    },
+    {
+      "epoch": 1.2723684210526316,
+      "grad_norm": 0.07872266626170443,
+      "learning_rate": 0.00014253913735832075,
+      "loss": 0.2642,
+      "step": 967
+    },
+    {
+      "epoch": 1.2736842105263158,
+      "grad_norm": 0.08023355760105522,
+      "learning_rate": 0.00014240053145816967,
+      "loss": 0.2568,
+      "step": 968
+    },
+    {
+      "epoch": 1.275,
+      "grad_norm": 0.0810413485993457,
+      "learning_rate": 0.00014226182617406996,
+      "loss": 0.2529,
+      "step": 969
+    },
+    {
+      "epoch": 1.2763157894736843,
+      "grad_norm": 0.08212814628024637,
+      "learning_rate": 0.00014212302183113732,
+      "loss": 0.2629,
+      "step": 970
+    },
+    {
+      "epoch": 1.2776315789473685,
+      "grad_norm": 0.08059136593229341,
+      "learning_rate": 0.00014198411875471955,
+      "loss": 0.2642,
+      "step": 971
+    },
+    {
+      "epoch": 1.2789473684210526,
+      "grad_norm": 0.08140236020647386,
+      "learning_rate": 0.00014184511727039612,
+      "loss": 0.2562,
+      "step": 972
+    },
+    {
+      "epoch": 1.2802631578947368,
+      "grad_norm": 0.08437007742270067,
+      "learning_rate": 0.00014170601770397692,
+      "loss": 0.2745,
+      "step": 973
+    },
+    {
+      "epoch": 1.2815789473684212,
+      "grad_norm": 0.08380982717641713,
+      "learning_rate": 0.00014156682038150183,
+      "loss": 0.2759,
+      "step": 974
+    },
+    {
+      "epoch": 1.2828947368421053,
+      "grad_norm": 0.0841772070431119,
+      "learning_rate": 0.00014142752562923988,
+      "loss": 0.2753,
+      "step": 975
+    },
+    {
+      "epoch": 1.2842105263157895,
+      "grad_norm": 0.08106806985375134,
+      "learning_rate": 0.0001412881337736885,
+      "loss": 0.2767,
+      "step": 976
+    },
+    {
+      "epoch": 1.2855263157894736,
+      "grad_norm": 0.0811005519720408,
+      "learning_rate": 0.00014114864514157258,
+      "loss": 0.2683,
+      "step": 977
+    },
+    {
+      "epoch": 1.2868421052631578,
+      "grad_norm": 0.08017428089067932,
+      "learning_rate": 0.00014100906005984403,
+      "loss": 0.2723,
+      "step": 978
+    },
+    {
+      "epoch": 1.2881578947368422,
+      "grad_norm": 0.07478870745109191,
+      "learning_rate": 0.00014086937885568067,
+      "loss": 0.2593,
+      "step": 979
+    },
+    {
+      "epoch": 1.2894736842105263,
+      "grad_norm": 0.07742432071526005,
+      "learning_rate": 0.00014072960185648577,
+      "loss": 0.2631,
+      "step": 980
+    },
+    {
+      "epoch": 1.2907894736842105,
+      "grad_norm": 0.07794666505896067,
+      "learning_rate": 0.000140589729389887,
+      "loss": 0.2644,
+      "step": 981
+    },
+    {
+      "epoch": 1.2921052631578949,
+      "grad_norm": 0.07946767921967093,
+      "learning_rate": 0.0001404497617837359,
+      "loss": 0.2609,
+      "step": 982
+    },
+    {
+      "epoch": 1.293421052631579,
+      "grad_norm": 0.08082944776271564,
+      "learning_rate": 0.00014030969936610697,
+      "loss": 0.2734,
+      "step": 983
+    },
+    {
+      "epoch": 1.2947368421052632,
+      "grad_norm": 0.08360749789540052,
+      "learning_rate": 0.00014016954246529696,
+      "loss": 0.263,
+      "step": 984
+    },
+    {
+      "epoch": 1.2960526315789473,
+      "grad_norm": 0.08375341936105894,
+      "learning_rate": 0.00014002929140982404,
+      "loss": 0.2639,
+      "step": 985
+    },
+    {
+      "epoch": 1.2973684210526315,
+      "grad_norm": 0.08336578005654308,
+      "learning_rate": 0.00013988894652842713,
+      "loss": 0.2777,
+      "step": 986
+    },
+    {
+      "epoch": 1.2986842105263157,
+      "grad_norm": 0.07859275037760449,
+      "learning_rate": 0.00013974850815006503,
+      "loss": 0.2609,
+      "step": 987
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.07972935132468843,
+      "learning_rate": 0.0001396079766039157,
+      "loss": 0.2593,
+      "step": 988
+    },
+    {
+      "epoch": 1.3013157894736842,
+      "grad_norm": 0.07852889738678417,
+      "learning_rate": 0.0001394673522193755,
+      "loss": 0.2651,
+      "step": 989
+    },
+    {
+      "epoch": 1.3026315789473684,
+      "grad_norm": 0.08131207293814138,
+      "learning_rate": 0.0001393266353260583,
+      "loss": 0.2636,
+      "step": 990
+    },
+    {
+      "epoch": 1.3039473684210527,
+      "grad_norm": 0.08085167622175171,
+      "learning_rate": 0.00013918582625379501,
+      "loss": 0.2581,
+      "step": 991
+    },
+    {
+      "epoch": 1.305263157894737,
+      "grad_norm": 0.08310768520116221,
+      "learning_rate": 0.00013904492533263244,
+      "loss": 0.2804,
+      "step": 992
+    },
+    {
+      "epoch": 1.306578947368421,
+      "grad_norm": 0.0835988231304127,
+      "learning_rate": 0.0001389039328928326,
+      "loss": 0.2764,
+      "step": 993
+    },
+    {
+      "epoch": 1.3078947368421052,
+      "grad_norm": 0.08043076516824947,
+      "learning_rate": 0.0001387628492648723,
+      "loss": 0.2643,
+      "step": 994
+    },
+    {
+      "epoch": 1.3092105263157894,
+      "grad_norm": 0.08124742205143656,
+      "learning_rate": 0.00013862167477944184,
+      "loss": 0.2779,
+      "step": 995
+    },
+    {
+      "epoch": 1.3105263157894738,
+      "grad_norm": 0.07855989514944883,
+      "learning_rate": 0.00013848040976744457,
+      "loss": 0.2641,
+      "step": 996
+    },
+    {
+      "epoch": 1.311842105263158,
+      "grad_norm": 0.08038296983596055,
+      "learning_rate": 0.00013833905455999603,
+      "loss": 0.2611,
+      "step": 997
+    },
+    {
+      "epoch": 1.313157894736842,
+      "grad_norm": 0.07754716363033522,
+      "learning_rate": 0.0001381976094884232,
+      "loss": 0.2565,
+      "step": 998
+    },
+    {
+      "epoch": 1.3144736842105262,
+      "grad_norm": 0.0771054502214774,
+      "learning_rate": 0.00013805607488426362,
+      "loss": 0.2667,
+      "step": 999
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.08092253741118172,
+      "learning_rate": 0.00013791445107926478,
+      "loss": 0.2715,
+      "step": 1000
+    },
+    {
+      "epoch": 1.3171052631578948,
+      "grad_norm": 0.08392681822232567,
+      "learning_rate": 0.00013777273840538312,
+      "loss": 0.2696,
+      "step": 1001
+    },
+    {
+      "epoch": 1.318421052631579,
+      "grad_norm": 0.08088919172679297,
+      "learning_rate": 0.00013763093719478358,
+      "loss": 0.262,
+      "step": 1002
+    },
+    {
+      "epoch": 1.319736842105263,
+      "grad_norm": 0.07966639118780608,
+      "learning_rate": 0.00013748904777983838,
+      "loss": 0.2632,
+      "step": 1003
+    },
+    {
+      "epoch": 1.3210526315789473,
+      "grad_norm": 0.07949592188633048,
+      "learning_rate": 0.00013734707049312673,
+      "loss": 0.2513,
+      "step": 1004
+    },
+    {
+      "epoch": 1.3223684210526316,
+      "grad_norm": 0.078682262346058,
+      "learning_rate": 0.00013720500566743362,
+      "loss": 0.2567,
+      "step": 1005
+    },
+    {
+      "epoch": 1.3236842105263158,
+      "grad_norm": 0.07844946772992165,
+      "learning_rate": 0.00013706285363574932,
+      "loss": 0.2743,
+      "step": 1006
+    },
+    {
+      "epoch": 1.325,
+      "grad_norm": 0.08305320278558118,
+      "learning_rate": 0.00013692061473126845,
+      "loss": 0.2623,
+      "step": 1007
+    },
+    {
+      "epoch": 1.3263157894736843,
+      "grad_norm": 0.08139900273296097,
+      "learning_rate": 0.00013677828928738934,
+      "loss": 0.269,
+      "step": 1008
+    },
+    {
+      "epoch": 1.3276315789473685,
+      "grad_norm": 0.0839606346260752,
+      "learning_rate": 0.00013663587763771304,
+      "loss": 0.2671,
+      "step": 1009
+    },
+    {
+      "epoch": 1.3289473684210527,
+      "grad_norm": 0.08150284952029307,
+      "learning_rate": 0.0001364933801160428,
+      "loss": 0.2568,
+      "step": 1010
+    },
+    {
+      "epoch": 1.3302631578947368,
+      "grad_norm": 0.07748260538941765,
+      "learning_rate": 0.00013635079705638298,
+      "loss": 0.2423,
+      "step": 1011
+    },
+    {
+      "epoch": 1.331578947368421,
+      "grad_norm": 0.07721123482181354,
+      "learning_rate": 0.00013620812879293863,
+      "loss": 0.2621,
+      "step": 1012
+    },
+    {
+      "epoch": 1.3328947368421051,
+      "grad_norm": 0.08010424676918745,
+      "learning_rate": 0.00013606537566011434,
+      "loss": 0.2643,
+      "step": 1013
+    },
+    {
+      "epoch": 1.3342105263157895,
+      "grad_norm": 0.07968708879421524,
+      "learning_rate": 0.00013592253799251376,
+      "loss": 0.2677,
+      "step": 1014
+    },
+    {
+      "epoch": 1.3355263157894737,
+      "grad_norm": 0.07865457628121457,
+      "learning_rate": 0.00013577961612493852,
+      "loss": 0.2646,
+      "step": 1015
+    },
+    {
+      "epoch": 1.3368421052631578,
+      "grad_norm": 0.07821958051935139,
+      "learning_rate": 0.00013563661039238785,
+      "loss": 0.2605,
+      "step": 1016
+    },
+    {
+      "epoch": 1.3381578947368422,
+      "grad_norm": 0.07970724490030985,
+      "learning_rate": 0.00013549352113005728,
+      "loss": 0.2591,
+      "step": 1017
+    },
+    {
+      "epoch": 1.3394736842105264,
+      "grad_norm": 0.08021459794844776,
+      "learning_rate": 0.00013535034867333837,
+      "loss": 0.2684,
+      "step": 1018
+    },
+    {
+      "epoch": 1.3407894736842105,
+      "grad_norm": 0.08161684018773531,
+      "learning_rate": 0.00013520709335781752,
+      "loss": 0.2579,
+      "step": 1019
+    },
+    {
+      "epoch": 1.3421052631578947,
+      "grad_norm": 0.0780970363121174,
+      "learning_rate": 0.00013506375551927547,
+      "loss": 0.258,
+      "step": 1020
+    },
+    {
+      "epoch": 1.3434210526315788,
+      "grad_norm": 0.08023973869702199,
+      "learning_rate": 0.00013492033549368618,
+      "loss": 0.2758,
+      "step": 1021
+    },
+    {
+      "epoch": 1.3447368421052632,
+      "grad_norm": 0.081261194750605,
+      "learning_rate": 0.00013477683361721657,
+      "loss": 0.2579,
+      "step": 1022
+    },
+    {
+      "epoch": 1.3460526315789474,
+      "grad_norm": 0.07810232344712505,
+      "learning_rate": 0.00013463325022622507,
+      "loss": 0.2597,
+      "step": 1023
+    },
+    {
+      "epoch": 1.3473684210526315,
+      "grad_norm": 0.07656812032411678,
+      "learning_rate": 0.00013448958565726144,
+      "loss": 0.2497,
+      "step": 1024
+    },
+    {
+      "epoch": 1.3486842105263157,
+      "grad_norm": 0.07962640765722634,
+      "learning_rate": 0.00013434584024706554,
+      "loss": 0.2703,
+      "step": 1025
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.07859791053686446,
+      "learning_rate": 0.00013420201433256689,
+      "loss": 0.2643,
+      "step": 1026
+    },
+    {
+      "epoch": 1.3513157894736842,
+      "grad_norm": 0.0785017433196338,
+      "learning_rate": 0.00013405810825088351,
+      "loss": 0.2554,
+      "step": 1027
+    },
+    {
+      "epoch": 1.3526315789473684,
+      "grad_norm": 0.07927097278835697,
+      "learning_rate": 0.00013391412233932149,
+      "loss": 0.2645,
+      "step": 1028
+    },
+    {
+      "epoch": 1.3539473684210526,
+      "grad_norm": 0.08028070891011724,
+      "learning_rate": 0.0001337700569353739,
+      "loss": 0.259,
+      "step": 1029
+    },
+    {
+      "epoch": 1.3552631578947367,
+      "grad_norm": 0.07903933780122868,
+      "learning_rate": 0.0001336259123767203,
+      "loss": 0.2687,
+      "step": 1030
+    },
+    {
+      "epoch": 1.356578947368421,
+      "grad_norm": 0.08789904311515143,
+      "learning_rate": 0.00013348168900122559,
+      "loss": 0.2653,
+      "step": 1031
+    },
+    {
+      "epoch": 1.3578947368421053,
+      "grad_norm": 0.0762050013831853,
+      "learning_rate": 0.00013333738714693956,
+      "loss": 0.2526,
+      "step": 1032
+    },
+    {
+      "epoch": 1.3592105263157894,
+      "grad_norm": 0.08083459753449757,
+      "learning_rate": 0.00013319300715209587,
+      "loss": 0.2687,
+      "step": 1033
+    },
+    {
+      "epoch": 1.3605263157894738,
+      "grad_norm": 0.0771191494993079,
+      "learning_rate": 0.00013304854935511143,
+      "loss": 0.2659,
+      "step": 1034
+    },
+    {
+      "epoch": 1.361842105263158,
+      "grad_norm": 0.0810612005537591,
+      "learning_rate": 0.00013290401409458532,
+      "loss": 0.2587,
+      "step": 1035
+    },
+    {
+      "epoch": 1.3631578947368421,
+      "grad_norm": 0.08077515208209074,
+      "learning_rate": 0.00013275940170929843,
+      "loss": 0.2763,
+      "step": 1036
+    },
+    {
+      "epoch": 1.3644736842105263,
+      "grad_norm": 0.0795071805895556,
+      "learning_rate": 0.00013261471253821227,
+      "loss": 0.2803,
+      "step": 1037
+    },
+    {
+      "epoch": 1.3657894736842104,
+      "grad_norm": 0.0783867946653792,
+      "learning_rate": 0.00013246994692046836,
+      "loss": 0.2659,
+      "step": 1038
+    },
+    {
+      "epoch": 1.3671052631578946,
+      "grad_norm": 0.07705812242711711,
+      "learning_rate": 0.0001323251051953874,
+      "loss": 0.2626,
+      "step": 1039
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.07981528649898714,
+      "learning_rate": 0.00013218018770246858,
+      "loss": 0.2652,
+      "step": 1040
+    },
+    {
+      "epoch": 1.3697368421052631,
+      "grad_norm": 0.07580835271521758,
+      "learning_rate": 0.00013203519478138852,
+      "loss": 0.2593,
+      "step": 1041
+    },
+    {
+      "epoch": 1.3710526315789473,
+      "grad_norm": 0.0786805131510188,
+      "learning_rate": 0.00013189012677200073,
+      "loss": 0.2583,
+      "step": 1042
+    },
+    {
+      "epoch": 1.3723684210526317,
+      "grad_norm": 0.07872461104976057,
+      "learning_rate": 0.00013174498401433474,
+      "loss": 0.2542,
+      "step": 1043
+    },
+    {
+      "epoch": 1.3736842105263158,
+      "grad_norm": 0.08393125286330598,
+      "learning_rate": 0.00013159976684859527,
+      "loss": 0.2717,
+      "step": 1044
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 0.08038226441935513,
+      "learning_rate": 0.00013145447561516138,
+      "loss": 0.2538,
+      "step": 1045
+    },
+    {
+      "epoch": 1.3763157894736842,
+      "grad_norm": 0.08104956663295539,
+      "learning_rate": 0.00013130911065458584,
+      "loss": 0.2666,
+      "step": 1046
+    },
+    {
+      "epoch": 1.3776315789473683,
+      "grad_norm": 0.08149965352071284,
+      "learning_rate": 0.00013116367230759415,
+      "loss": 0.2741,
+      "step": 1047
+    },
+    {
+      "epoch": 1.3789473684210527,
+      "grad_norm": 0.0800773120109665,
+      "learning_rate": 0.00013101816091508388,
+      "loss": 0.2621,
+      "step": 1048
+    },
+    {
+      "epoch": 1.3802631578947369,
+      "grad_norm": 0.07883317602840334,
+      "learning_rate": 0.00013087257681812376,
+      "loss": 0.2627,
+      "step": 1049
+    },
+    {
+      "epoch": 1.381578947368421,
+      "grad_norm": 0.07808540918916572,
+      "learning_rate": 0.00013072692035795305,
+      "loss": 0.2661,
+      "step": 1050
+    },
+    {
+      "epoch": 1.3828947368421054,
+      "grad_norm": 0.07784582555862558,
+      "learning_rate": 0.00013058119187598047,
+      "loss": 0.2605,
+      "step": 1051
+    },
+    {
+      "epoch": 1.3842105263157896,
+      "grad_norm": 0.0789115552020285,
+      "learning_rate": 0.0001304353917137836,
+      "loss": 0.2591,
+      "step": 1052
+    },
+    {
+      "epoch": 1.3855263157894737,
+      "grad_norm": 0.07707043429238686,
+      "learning_rate": 0.00013028952021310812,
+      "loss": 0.2602,
+      "step": 1053
+    },
+    {
+      "epoch": 1.3868421052631579,
+      "grad_norm": 0.0764917703073864,
+      "learning_rate": 0.00013014357771586686,
+      "loss": 0.252,
+      "step": 1054
+    },
+    {
+      "epoch": 1.388157894736842,
+      "grad_norm": 0.08152267179709696,
+      "learning_rate": 0.000129997564564139,
+      "loss": 0.2761,
+      "step": 1055
+    },
+    {
+      "epoch": 1.3894736842105262,
+      "grad_norm": 0.07488321865899004,
+      "learning_rate": 0.00012985148110016947,
+      "loss": 0.2557,
+      "step": 1056
+    },
+    {
+      "epoch": 1.3907894736842106,
+      "grad_norm": 0.08223478003170398,
+      "learning_rate": 0.00012970532766636787,
+      "loss": 0.2758,
+      "step": 1057
+    },
+    {
+      "epoch": 1.3921052631578947,
+      "grad_norm": 0.08146309512224036,
+      "learning_rate": 0.00012955910460530788,
+      "loss": 0.2631,
+      "step": 1058
+    },
+    {
+      "epoch": 1.393421052631579,
+      "grad_norm": 0.08083129097727351,
+      "learning_rate": 0.00012941281225972636,
+      "loss": 0.2848,
+      "step": 1059
+    },
+    {
+      "epoch": 1.3947368421052633,
+      "grad_norm": 0.07757651778516991,
+      "learning_rate": 0.0001292664509725226,
+      "loss": 0.2524,
+      "step": 1060
+    },
+    {
+      "epoch": 1.3960526315789474,
+      "grad_norm": 0.0763538974448064,
+      "learning_rate": 0.0001291200210867574,
+      "loss": 0.257,
+      "step": 1061
+    },
+    {
+      "epoch": 1.3973684210526316,
+      "grad_norm": 0.07882164295488549,
+      "learning_rate": 0.0001289735229456525,
+      "loss": 0.2687,
+      "step": 1062
+    },
+    {
+      "epoch": 1.3986842105263158,
+      "grad_norm": 0.07892589978626854,
+      "learning_rate": 0.0001288269568925894,
+      "loss": 0.2704,
+      "step": 1063
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.07820774930633734,
+      "learning_rate": 0.00012868032327110904,
+      "loss": 0.2619,
+      "step": 1064
+    },
+    {
+      "epoch": 1.4013157894736843,
+      "grad_norm": 0.07815937274547557,
+      "learning_rate": 0.00012853362242491053,
+      "loss": 0.2717,
+      "step": 1065
+    },
+    {
+      "epoch": 1.4026315789473685,
+      "grad_norm": 0.07522447260364416,
+      "learning_rate": 0.0001283868546978507,
+      "loss": 0.2554,
+      "step": 1066
+    },
+    {
+      "epoch": 1.4039473684210526,
+      "grad_norm": 0.08070192461804411,
+      "learning_rate": 0.00012824002043394298,
+      "loss": 0.2754,
+      "step": 1067
+    },
+    {
+      "epoch": 1.4052631578947368,
+      "grad_norm": 0.0803973317378383,
+      "learning_rate": 0.00012809311997735696,
+      "loss": 0.2766,
+      "step": 1068
+    },
+    {
+      "epoch": 1.4065789473684212,
+      "grad_norm": 0.07866656181841848,
+      "learning_rate": 0.00012794615367241717,
+      "loss": 0.2668,
+      "step": 1069
+    },
+    {
+      "epoch": 1.4078947368421053,
+      "grad_norm": 0.0794268932038752,
+      "learning_rate": 0.00012779912186360268,
+      "loss": 0.2675,
+      "step": 1070
+    },
+    {
+      "epoch": 1.4092105263157895,
+      "grad_norm": 0.07900705819829199,
+      "learning_rate": 0.0001276520248955459,
+      "loss": 0.2635,
+      "step": 1071
+    },
+    {
+      "epoch": 1.4105263157894736,
+      "grad_norm": 0.07918211928299614,
+      "learning_rate": 0.00012750486311303218,
+      "loss": 0.262,
+      "step": 1072
+    },
+    {
+      "epoch": 1.4118421052631578,
+      "grad_norm": 0.0819965398370594,
+      "learning_rate": 0.00012735763686099862,
+      "loss": 0.2602,
+      "step": 1073
+    },
+    {
+      "epoch": 1.4131578947368422,
+      "grad_norm": 0.07831915554583863,
+      "learning_rate": 0.00012721034648453353,
+      "loss": 0.2628,
+      "step": 1074
+    },
+    {
+      "epoch": 1.4144736842105263,
+      "grad_norm": 0.07786089842983143,
+      "learning_rate": 0.00012706299232887543,
+      "loss": 0.2561,
+      "step": 1075
+    },
+    {
+      "epoch": 1.4157894736842105,
+      "grad_norm": 0.07971967408570595,
+      "learning_rate": 0.00012691557473941243,
+      "loss": 0.2656,
+      "step": 1076
+    },
+    {
+      "epoch": 1.4171052631578949,
+      "grad_norm": 0.07801118181055779,
+      "learning_rate": 0.00012676809406168133,
+      "loss": 0.2652,
+      "step": 1077
+    },
+    {
+      "epoch": 1.418421052631579,
+      "grad_norm": 0.08515813484361862,
+      "learning_rate": 0.00012662055064136668,
+      "loss": 0.2753,
+      "step": 1078
+    },
+    {
+      "epoch": 1.4197368421052632,
+      "grad_norm": 0.08068756136226636,
+      "learning_rate": 0.00012647294482430024,
+      "loss": 0.2634,
+      "step": 1079
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.0794228149168458,
+      "learning_rate": 0.00012632527695645993,
+      "loss": 0.2613,
+      "step": 1080
+    },
+    {
+      "epoch": 1.4223684210526315,
+      "grad_norm": 0.07796131697715634,
+      "learning_rate": 0.00012617754738396915,
+      "loss": 0.2665,
+      "step": 1081
+    },
+    {
+      "epoch": 1.4236842105263157,
+      "grad_norm": 0.07971630587719376,
+      "learning_rate": 0.00012602975645309593,
+      "loss": 0.2631,
+      "step": 1082
+    },
+    {
+      "epoch": 1.425,
+      "grad_norm": 0.07712944683367323,
+      "learning_rate": 0.00012588190451025207,
+      "loss": 0.2509,
+      "step": 1083
+    },
+    {
+      "epoch": 1.4263157894736842,
+      "grad_norm": 0.08097034232881713,
+      "learning_rate": 0.0001257339919019925,
+      "loss": 0.2649,
+      "step": 1084
+    },
+    {
+      "epoch": 1.4276315789473684,
+      "grad_norm": 0.07618596280882184,
+      "learning_rate": 0.0001255860189750142,
+      "loss": 0.2613,
+      "step": 1085
+    },
+    {
+      "epoch": 1.4289473684210527,
+      "grad_norm": 0.08305767349447553,
+      "learning_rate": 0.00012543798607615565,
+      "loss": 0.2758,
+      "step": 1086
+    },
+    {
+      "epoch": 1.430263157894737,
+      "grad_norm": 0.08146614433553916,
+      "learning_rate": 0.0001252898935523958,
+      "loss": 0.2819,
+      "step": 1087
+    },
+    {
+      "epoch": 1.431578947368421,
+      "grad_norm": 0.07805110779507855,
+      "learning_rate": 0.00012514174175085345,
+      "loss": 0.2649,
+      "step": 1088
+    },
+    {
+      "epoch": 1.4328947368421052,
+      "grad_norm": 0.07969018389253195,
+      "learning_rate": 0.0001249935310187863,
+      "loss": 0.2637,
+      "step": 1089
+    },
+    {
+      "epoch": 1.4342105263157894,
+      "grad_norm": 0.07956536699437335,
+      "learning_rate": 0.00012484526170359012,
+      "loss": 0.2651,
+      "step": 1090
+    },
+    {
+      "epoch": 1.4355263157894738,
+      "grad_norm": 0.07836456115097726,
+      "learning_rate": 0.0001246969341527981,
+      "loss": 0.2645,
+      "step": 1091
+    },
+    {
+      "epoch": 1.436842105263158,
+      "grad_norm": 0.07976260454428451,
+      "learning_rate": 0.00012454854871407994,
+      "loss": 0.2824,
+      "step": 1092
+    },
+    {
+      "epoch": 1.438157894736842,
+      "grad_norm": 0.07726676671750074,
+      "learning_rate": 0.00012440010573524086,
+      "loss": 0.2632,
+      "step": 1093
+    },
+    {
+      "epoch": 1.4394736842105262,
+      "grad_norm": 0.07821021819439651,
+      "learning_rate": 0.00012425160556422114,
+      "loss": 0.2731,
+      "step": 1094
+    },
+    {
+      "epoch": 1.4407894736842106,
+      "grad_norm": 0.08013778051504493,
+      "learning_rate": 0.00012410304854909495,
+      "loss": 0.2559,
+      "step": 1095
+    },
+    {
+      "epoch": 1.4421052631578948,
+      "grad_norm": 0.08401500576489536,
+      "learning_rate": 0.0001239544350380699,
+      "loss": 0.2719,
+      "step": 1096
+    },
+    {
+      "epoch": 1.443421052631579,
+      "grad_norm": 0.08434562563421157,
+      "learning_rate": 0.0001238057653794858,
+      "loss": 0.269,
+      "step": 1097
+    },
+    {
+      "epoch": 1.444736842105263,
+      "grad_norm": 0.07770950324688412,
+      "learning_rate": 0.00012365703992181425,
+      "loss": 0.2623,
+      "step": 1098
+    },
+    {
+      "epoch": 1.4460526315789473,
+      "grad_norm": 0.08224554301801708,
+      "learning_rate": 0.0001235082590136575,
+      "loss": 0.2808,
+      "step": 1099
+    },
+    {
+      "epoch": 1.4473684210526316,
+      "grad_norm": 0.07921911039298107,
+      "learning_rate": 0.00012335942300374788,
+      "loss": 0.2598,
+      "step": 1100
+    },
+    {
+      "epoch": 1.4486842105263158,
+      "grad_norm": 0.08061258248097088,
+      "learning_rate": 0.0001232105322409468,
+      "loss": 0.2674,
+      "step": 1101
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.0777388780233582,
+      "learning_rate": 0.00012306158707424403,
+      "loss": 0.2498,
+      "step": 1102
+    },
+    {
+      "epoch": 1.4513157894736843,
+      "grad_norm": 0.07794347014147819,
+      "learning_rate": 0.00012291258785275687,
+      "loss": 0.2569,
+      "step": 1103
+    },
+    {
+      "epoch": 1.4526315789473685,
+      "grad_norm": 0.08178650020126337,
+      "learning_rate": 0.00012276353492572935,
+      "loss": 0.2798,
+      "step": 1104
+    },
+    {
+      "epoch": 1.4539473684210527,
+      "grad_norm": 0.08037719214824916,
+      "learning_rate": 0.0001226144286425313,
+      "loss": 0.2684,
+      "step": 1105
+    },
+    {
+      "epoch": 1.4552631578947368,
+      "grad_norm": 0.0763595502941657,
+      "learning_rate": 0.00012246526935265768,
+      "loss": 0.2483,
+      "step": 1106
+    },
+    {
+      "epoch": 1.456578947368421,
+      "grad_norm": 0.07688959299684457,
+      "learning_rate": 0.00012231605740572766,
+      "loss": 0.2524,
+      "step": 1107
+    },
+    {
+      "epoch": 1.4578947368421051,
+      "grad_norm": 0.07894151742409149,
+      "learning_rate": 0.00012216679315148386,
+      "loss": 0.2642,
+      "step": 1108
+    },
+    {
+      "epoch": 1.4592105263157895,
+      "grad_norm": 0.07777205049498395,
+      "learning_rate": 0.00012201747693979151,
+      "loss": 0.2499,
+      "step": 1109
+    },
+    {
+      "epoch": 1.4605263157894737,
+      "grad_norm": 0.08207208250132972,
+      "learning_rate": 0.0001218681091206376,
+      "loss": 0.2751,
+      "step": 1110
+    },
+    {
+      "epoch": 1.4618421052631578,
+      "grad_norm": 0.07844860347132009,
+      "learning_rate": 0.00012171869004413006,
+      "loss": 0.2602,
+      "step": 1111
+    },
+    {
+      "epoch": 1.4631578947368422,
+      "grad_norm": 0.07856698522588018,
+      "learning_rate": 0.00012156922006049702,
+      "loss": 0.2658,
+      "step": 1112
+    },
+    {
+      "epoch": 1.4644736842105264,
+      "grad_norm": 0.07684231974080626,
+      "learning_rate": 0.00012141969952008591,
+      "loss": 0.2519,
+      "step": 1113
+    },
+    {
+      "epoch": 1.4657894736842105,
+      "grad_norm": 0.07956756231471081,
+      "learning_rate": 0.00012127012877336266,
+      "loss": 0.2587,
+      "step": 1114
+    },
+    {
+      "epoch": 1.4671052631578947,
+      "grad_norm": 0.07797778646499881,
+      "learning_rate": 0.00012112050817091087,
+      "loss": 0.2435,
+      "step": 1115
+    },
+    {
+      "epoch": 1.4684210526315788,
+      "grad_norm": 0.07789957155018702,
+      "learning_rate": 0.00012097083806343103,
+      "loss": 0.2675,
+      "step": 1116
+    },
+    {
+      "epoch": 1.4697368421052632,
+      "grad_norm": 0.08090700227859492,
+      "learning_rate": 0.00012082111880173965,
+      "loss": 0.2772,
+      "step": 1117
+    },
+    {
+      "epoch": 1.4710526315789474,
+      "grad_norm": 0.07661313410176111,
+      "learning_rate": 0.0001206713507367684,
+      "loss": 0.2571,
+      "step": 1118
+    },
+    {
+      "epoch": 1.4723684210526315,
+      "grad_norm": 0.07907971896935535,
+      "learning_rate": 0.00012052153421956342,
+      "loss": 0.2649,
+      "step": 1119
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.07996349059046486,
+      "learning_rate": 0.00012037166960128443,
+      "loss": 0.2536,
+      "step": 1120
+    },
+    {
+      "epoch": 1.475,
+      "grad_norm": 0.08235866500979575,
+      "learning_rate": 0.00012022175723320381,
+      "loss": 0.2778,
+      "step": 1121
+    },
+    {
+      "epoch": 1.4763157894736842,
+      "grad_norm": 0.08262772185376702,
+      "learning_rate": 0.00012007179746670592,
+      "loss": 0.2717,
+      "step": 1122
+    },
+    {
+      "epoch": 1.4776315789473684,
+      "grad_norm": 0.07931030156939642,
+      "learning_rate": 0.00011992179065328618,
+      "loss": 0.2556,
+      "step": 1123
+    },
+    {
+      "epoch": 1.4789473684210526,
+      "grad_norm": 0.0818492117725006,
+      "learning_rate": 0.00011977173714455034,
+      "loss": 0.2734,
+      "step": 1124
+    },
+    {
+      "epoch": 1.4802631578947367,
+      "grad_norm": 0.08007910950748007,
+      "learning_rate": 0.0001196216372922136,
+      "loss": 0.2591,
+      "step": 1125
+    },
+    {
+      "epoch": 1.481578947368421,
+      "grad_norm": 0.07912376020739983,
+      "learning_rate": 0.00011947149144809969,
+      "loss": 0.2573,
+      "step": 1126
+    },
+    {
+      "epoch": 1.4828947368421053,
+      "grad_norm": 0.07919295195700186,
+      "learning_rate": 0.00011932129996414023,
+      "loss": 0.2796,
+      "step": 1127
+    },
+    {
+      "epoch": 1.4842105263157894,
+      "grad_norm": 0.0768053198795628,
+      "learning_rate": 0.00011917106319237386,
+      "loss": 0.258,
+      "step": 1128
+    },
+    {
+      "epoch": 1.4855263157894738,
+      "grad_norm": 0.0751828643806364,
+      "learning_rate": 0.00011902078148494525,
+      "loss": 0.2482,
+      "step": 1129
+    },
+    {
+      "epoch": 1.486842105263158,
+      "grad_norm": 0.07573342946120709,
+      "learning_rate": 0.00011887045519410442,
+      "loss": 0.2515,
+      "step": 1130
+    },
+    {
+      "epoch": 1.4881578947368421,
+      "grad_norm": 0.08033169720613252,
+      "learning_rate": 0.00011872008467220599,
+      "loss": 0.2727,
+      "step": 1131
+    },
+    {
+      "epoch": 1.4894736842105263,
+      "grad_norm": 0.07910729035555242,
+      "learning_rate": 0.00011856967027170818,
+      "loss": 0.2753,
+      "step": 1132
+    },
+    {
+      "epoch": 1.4907894736842104,
+      "grad_norm": 0.07820164568049928,
+      "learning_rate": 0.00011841921234517206,
+      "loss": 0.2612,
+      "step": 1133
+    },
+    {
+      "epoch": 1.4921052631578946,
+      "grad_norm": 0.07934456848043618,
+      "learning_rate": 0.00011826871124526071,
+      "loss": 0.2619,
+      "step": 1134
+    },
+    {
+      "epoch": 1.493421052631579,
+      "grad_norm": 0.08088874374294254,
+      "learning_rate": 0.00011811816732473841,
+      "loss": 0.2582,
+      "step": 1135
+    },
+    {
+      "epoch": 1.4947368421052631,
+      "grad_norm": 0.07877981008540878,
+      "learning_rate": 0.00011796758093646989,
+      "loss": 0.2548,
+      "step": 1136
+    },
+    {
+      "epoch": 1.4960526315789473,
+      "grad_norm": 0.07864927828591978,
+      "learning_rate": 0.00011781695243341932,
+      "loss": 0.2587,
+      "step": 1137
+    },
+    {
+      "epoch": 1.4973684210526317,
+      "grad_norm": 0.08496699143061613,
+      "learning_rate": 0.0001176662821686496,
+      "loss": 0.2769,
+      "step": 1138
+    },
+    {
+      "epoch": 1.4986842105263158,
+      "grad_norm": 0.07790722036110928,
+      "learning_rate": 0.00011751557049532153,
+      "loss": 0.2625,
+      "step": 1139
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.07732825022547576,
+      "learning_rate": 0.00011736481776669306,
+      "loss": 0.2595,
+      "step": 1140
+    },
+    {
+      "epoch": 1.5013157894736842,
+      "grad_norm": 0.07839345817113182,
+      "learning_rate": 0.00011721402433611818,
+      "loss": 0.264,
+      "step": 1141
+    },
+    {
+      "epoch": 1.5026315789473683,
+      "grad_norm": 0.0797985656438268,
+      "learning_rate": 0.00011706319055704642,
+      "loss": 0.2762,
+      "step": 1142
+    },
+    {
+      "epoch": 1.5039473684210525,
+      "grad_norm": 0.07859356519091114,
+      "learning_rate": 0.00011691231678302187,
+      "loss": 0.2589,
+      "step": 1143
+    },
+    {
+      "epoch": 1.5052631578947369,
+      "grad_norm": 0.08132159915933083,
+      "learning_rate": 0.00011676140336768236,
+      "loss": 0.2676,
+      "step": 1144
+    },
+    {
+      "epoch": 1.506578947368421,
+      "grad_norm": 0.07821055809299678,
+      "learning_rate": 0.0001166104506647586,
+      "loss": 0.2647,
+      "step": 1145
+    },
+    {
+      "epoch": 1.5078947368421054,
+      "grad_norm": 0.07762701396147464,
+      "learning_rate": 0.00011645945902807341,
+      "loss": 0.2646,
+      "step": 1146
+    },
+    {
+      "epoch": 1.5092105263157896,
+      "grad_norm": 0.07496175871914876,
+      "learning_rate": 0.00011630842881154085,
+      "loss": 0.2576,
+      "step": 1147
+    },
+    {
+      "epoch": 1.5105263157894737,
+      "grad_norm": 0.0784666282944514,
+      "learning_rate": 0.00011615736036916549,
+      "loss": 0.2679,
+      "step": 1148
+    },
+    {
+      "epoch": 1.5118421052631579,
+      "grad_norm": 0.07722067192546046,
+      "learning_rate": 0.0001160062540550414,
+      "loss": 0.2639,
+      "step": 1149
+    },
+    {
+      "epoch": 1.513157894736842,
+      "grad_norm": 0.08059485240498777,
+      "learning_rate": 0.00011585511022335142,
+      "loss": 0.2764,
+      "step": 1150
+    },
+    {
+      "epoch": 1.5144736842105262,
+      "grad_norm": 0.08129025875363483,
+      "learning_rate": 0.00011570392922836644,
+      "loss": 0.2714,
+      "step": 1151
+    },
+    {
+      "epoch": 1.5157894736842106,
+      "grad_norm": 0.07486275958049285,
+      "learning_rate": 0.00011555271142444433,
+      "loss": 0.2517,
+      "step": 1152
+    },
+    {
+      "epoch": 1.5171052631578947,
+      "grad_norm": 0.08143946348918624,
+      "learning_rate": 0.0001154014571660293,
+      "loss": 0.2747,
+      "step": 1153
+    },
+    {
+      "epoch": 1.518421052631579,
+      "grad_norm": 0.07923671682954142,
+      "learning_rate": 0.00011525016680765102,
+      "loss": 0.2485,
+      "step": 1154
+    },
+    {
+      "epoch": 1.5197368421052633,
+      "grad_norm": 0.08011623724583099,
+      "learning_rate": 0.00011509884070392369,
+      "loss": 0.2656,
+      "step": 1155
+    },
+    {
+      "epoch": 1.5210526315789474,
+      "grad_norm": 0.07946415372441243,
+      "learning_rate": 0.00011494747920954545,
+      "loss": 0.2572,
+      "step": 1156
+    },
+    {
+      "epoch": 1.5223684210526316,
+      "grad_norm": 0.0812375151138501,
+      "learning_rate": 0.00011479608267929722,
+      "loss": 0.2697,
+      "step": 1157
+    },
+    {
+      "epoch": 1.5236842105263158,
+      "grad_norm": 0.07578132261810436,
+      "learning_rate": 0.00011464465146804217,
+      "loss": 0.2411,
+      "step": 1158
+    },
+    {
+      "epoch": 1.525,
+      "grad_norm": 0.07784132005694626,
+      "learning_rate": 0.00011449318593072466,
+      "loss": 0.2678,
+      "step": 1159
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.07909385203173394,
+      "learning_rate": 0.00011434168642236964,
+      "loss": 0.2754,
+      "step": 1160
+    },
+    {
+      "epoch": 1.5276315789473685,
+      "grad_norm": 0.0776880493894959,
+      "learning_rate": 0.00011419015329808157,
+      "loss": 0.2655,
+      "step": 1161
+    },
+    {
+      "epoch": 1.5289473684210526,
+      "grad_norm": 0.07765848299877728,
+      "learning_rate": 0.00011403858691304373,
+      "loss": 0.2607,
+      "step": 1162
+    },
+    {
+      "epoch": 1.530263157894737,
+      "grad_norm": 0.07646804543626935,
+      "learning_rate": 0.00011388698762251732,
+      "loss": 0.2596,
+      "step": 1163
+    },
+    {
+      "epoch": 1.5315789473684212,
+      "grad_norm": 0.07740933938489439,
+      "learning_rate": 0.00011373535578184082,
+      "loss": 0.2707,
+      "step": 1164
+    },
+    {
+      "epoch": 1.5328947368421053,
+      "grad_norm": 0.07996711741187337,
+      "learning_rate": 0.00011358369174642887,
+      "loss": 0.2664,
+      "step": 1165
+    },
+    {
+      "epoch": 1.5342105263157895,
+      "grad_norm": 0.07987614614098584,
+      "learning_rate": 0.00011343199587177155,
+      "loss": 0.2714,
+      "step": 1166
+    },
+    {
+      "epoch": 1.5355263157894736,
+      "grad_norm": 0.07887783815136196,
+      "learning_rate": 0.00011328026851343367,
+      "loss": 0.2548,
+      "step": 1167
+    },
+    {
+      "epoch": 1.5368421052631578,
+      "grad_norm": 0.07775100523637381,
+      "learning_rate": 0.00011312851002705383,
+      "loss": 0.2587,
+      "step": 1168
+    },
+    {
+      "epoch": 1.538157894736842,
+      "grad_norm": 0.08179602367804785,
+      "learning_rate": 0.00011297672076834348,
+      "loss": 0.2743,
+      "step": 1169
+    },
+    {
+      "epoch": 1.5394736842105263,
+      "grad_norm": 0.08166670313368278,
+      "learning_rate": 0.00011282490109308633,
+      "loss": 0.2743,
+      "step": 1170
+    },
+    {
+      "epoch": 1.5407894736842105,
+      "grad_norm": 0.07926586858824067,
+      "learning_rate": 0.00011267305135713726,
+      "loss": 0.2532,
+      "step": 1171
+    },
+    {
+      "epoch": 1.5421052631578949,
+      "grad_norm": 0.07935859462462387,
+      "learning_rate": 0.00011252117191642175,
+      "loss": 0.2586,
+      "step": 1172
+    },
+    {
+      "epoch": 1.543421052631579,
+      "grad_norm": 0.07771700777076371,
+      "learning_rate": 0.00011236926312693479,
+      "loss": 0.2556,
+      "step": 1173
+    },
+    {
+      "epoch": 1.5447368421052632,
+      "grad_norm": 0.07944739071127381,
+      "learning_rate": 0.00011221732534474019,
+      "loss": 0.2669,
+      "step": 1174
+    },
+    {
+      "epoch": 1.5460526315789473,
+      "grad_norm": 0.07926172086790113,
+      "learning_rate": 0.00011206535892596975,
+      "loss": 0.2578,
+      "step": 1175
+    },
+    {
+      "epoch": 1.5473684210526315,
+      "grad_norm": 0.07836482244209485,
+      "learning_rate": 0.00011191336422682237,
+      "loss": 0.266,
+      "step": 1176
+    },
+    {
+      "epoch": 1.5486842105263157,
+      "grad_norm": 0.08060052790786207,
+      "learning_rate": 0.00011176134160356327,
+      "loss": 0.27,
+      "step": 1177
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.07857242666783919,
+      "learning_rate": 0.00011160929141252303,
+      "loss": 0.2561,
+      "step": 1178
+    },
+    {
+      "epoch": 1.5513157894736842,
+      "grad_norm": 0.0762732065222906,
+      "learning_rate": 0.00011145721401009694,
+      "loss": 0.265,
+      "step": 1179
+    },
+    {
+      "epoch": 1.5526315789473686,
+      "grad_norm": 0.07588799278326497,
+      "learning_rate": 0.00011130510975274409,
+      "loss": 0.2659,
+      "step": 1180
+    },
+    {
+      "epoch": 1.5539473684210527,
+      "grad_norm": 0.0782353560349947,
+      "learning_rate": 0.00011115297899698638,
+      "loss": 0.2622,
+      "step": 1181
+    },
+    {
+      "epoch": 1.555263157894737,
+      "grad_norm": 0.0779137347020748,
+      "learning_rate": 0.00011100082209940795,
+      "loss": 0.2581,
+      "step": 1182
+    },
+    {
+      "epoch": 1.556578947368421,
+      "grad_norm": 0.07626200398110206,
+      "learning_rate": 0.00011084863941665415,
+      "loss": 0.2525,
+      "step": 1183
+    },
+    {
+      "epoch": 1.5578947368421052,
+      "grad_norm": 0.08075601397948919,
+      "learning_rate": 0.00011069643130543084,
+      "loss": 0.2589,
+      "step": 1184
+    },
+    {
+      "epoch": 1.5592105263157894,
+      "grad_norm": 0.07890719633469666,
+      "learning_rate": 0.00011054419812250338,
+      "loss": 0.2609,
+      "step": 1185
+    },
+    {
+      "epoch": 1.5605263157894735,
+      "grad_norm": 0.0789865706647385,
+      "learning_rate": 0.00011039194022469597,
+      "loss": 0.2515,
+      "step": 1186
+    },
+    {
+      "epoch": 1.561842105263158,
+      "grad_norm": 0.0796846545034331,
+      "learning_rate": 0.0001102396579688907,
+      "loss": 0.2687,
+      "step": 1187
+    },
+    {
+      "epoch": 1.563157894736842,
+      "grad_norm": 0.07938699909977037,
+      "learning_rate": 0.00011008735171202684,
+      "loss": 0.2636,
+      "step": 1188
+    },
+    {
+      "epoch": 1.5644736842105265,
+      "grad_norm": 0.0810811457705607,
+      "learning_rate": 0.00010993502181109978,
+      "loss": 0.2544,
+      "step": 1189
+    },
+    {
+      "epoch": 1.5657894736842106,
+      "grad_norm": 0.0811343931888347,
+      "learning_rate": 0.0001097826686231604,
+      "loss": 0.274,
+      "step": 1190
+    },
+    {
+      "epoch": 1.5671052631578948,
+      "grad_norm": 0.0786799724878539,
+      "learning_rate": 0.00010963029250531418,
+      "loss": 0.2619,
+      "step": 1191
+    },
+    {
+      "epoch": 1.568421052631579,
+      "grad_norm": 0.07801305259558834,
+      "learning_rate": 0.00010947789381472035,
+      "loss": 0.2575,
+      "step": 1192
+    },
+    {
+      "epoch": 1.569736842105263,
+      "grad_norm": 0.08019971785855351,
+      "learning_rate": 0.00010932547290859103,
+      "loss": 0.2592,
+      "step": 1193
+    },
+    {
+      "epoch": 1.5710526315789473,
+      "grad_norm": 0.07693908653004418,
+      "learning_rate": 0.00010917303014419036,
+      "loss": 0.2665,
+      "step": 1194
+    },
+    {
+      "epoch": 1.5723684210526314,
+      "grad_norm": 0.07646993423396128,
+      "learning_rate": 0.00010902056587883378,
+      "loss": 0.245,
+      "step": 1195
+    },
+    {
+      "epoch": 1.5736842105263158,
+      "grad_norm": 0.080428348690664,
+      "learning_rate": 0.00010886808046988717,
+      "loss": 0.2588,
+      "step": 1196
+    },
+    {
+      "epoch": 1.575,
+      "grad_norm": 0.07877667825439123,
+      "learning_rate": 0.00010871557427476583,
+      "loss": 0.2688,
+      "step": 1197
+    },
+    {
+      "epoch": 1.5763157894736843,
+      "grad_norm": 0.07769524509043618,
+      "learning_rate": 0.0001085630476509339,
+      "loss": 0.2588,
+      "step": 1198
+    },
+    {
+      "epoch": 1.5776315789473685,
+      "grad_norm": 0.08089328257237792,
+      "learning_rate": 0.00010841050095590335,
+      "loss": 0.2643,
+      "step": 1199
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.07938294687717415,
+      "learning_rate": 0.00010825793454723325,
+      "loss": 0.2678,
+      "step": 1200
+    },
+    {
+      "epoch": 1.5802631578947368,
+      "grad_norm": 0.07853030205890081,
+      "learning_rate": 0.0001081053487825288,
+      "loss": 0.2606,
+      "step": 1201
+    },
+    {
+      "epoch": 1.581578947368421,
+      "grad_norm": 0.07774175526579997,
+      "learning_rate": 0.00010795274401944058,
+      "loss": 0.2582,
+      "step": 1202
+    },
+    {
+      "epoch": 1.5828947368421051,
+      "grad_norm": 0.07578431858186546,
+      "learning_rate": 0.00010780012061566378,
+      "loss": 0.2568,
+      "step": 1203
+    },
+    {
+      "epoch": 1.5842105263157895,
+      "grad_norm": 0.07606000180404633,
+      "learning_rate": 0.00010764747892893723,
+      "loss": 0.2646,
+      "step": 1204
+    },
+    {
+      "epoch": 1.5855263157894737,
+      "grad_norm": 0.07591132081458318,
+      "learning_rate": 0.0001074948193170426,
+      "loss": 0.2517,
+      "step": 1205
+    },
+    {
+      "epoch": 1.586842105263158,
+      "grad_norm": 0.0781464844924248,
+      "learning_rate": 0.00010734214213780354,
+      "loss": 0.2491,
+      "step": 1206
+    },
+    {
+      "epoch": 1.5881578947368422,
+      "grad_norm": 0.07902701850224841,
+      "learning_rate": 0.000107189447749085,
+      "loss": 0.2653,
+      "step": 1207
+    },
+    {
+      "epoch": 1.5894736842105264,
+      "grad_norm": 0.08084483037179403,
+      "learning_rate": 0.00010703673650879218,
+      "loss": 0.2625,
+      "step": 1208
+    },
+    {
+      "epoch": 1.5907894736842105,
+      "grad_norm": 0.0773678893977578,
+      "learning_rate": 0.00010688400877486978,
+      "loss": 0.2563,
+      "step": 1209
+    },
+    {
+      "epoch": 1.5921052631578947,
+      "grad_norm": 0.07876897624596318,
+      "learning_rate": 0.00010673126490530112,
+      "loss": 0.2649,
+      "step": 1210
+    },
+    {
+      "epoch": 1.5934210526315788,
+      "grad_norm": 0.0804198373742344,
+      "learning_rate": 0.00010657850525810748,
+      "loss": 0.268,
+      "step": 1211
+    },
+    {
+      "epoch": 1.594736842105263,
+      "grad_norm": 0.08109356740988712,
+      "learning_rate": 0.00010642573019134703,
+      "loss": 0.2634,
+      "step": 1212
+    },
+    {
+      "epoch": 1.5960526315789474,
+      "grad_norm": 0.08129604963653887,
+      "learning_rate": 0.00010627294006311404,
+      "loss": 0.2852,
+      "step": 1213
+    },
+    {
+      "epoch": 1.5973684210526315,
+      "grad_norm": 0.07643522766740891,
+      "learning_rate": 0.00010612013523153812,
+      "loss": 0.261,
+      "step": 1214
+    },
+    {
+      "epoch": 1.598684210526316,
+      "grad_norm": 0.07695348511929585,
+      "learning_rate": 0.0001059673160547834,
+      "loss": 0.2651,
+      "step": 1215
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.07630900188908298,
+      "learning_rate": 0.00010581448289104758,
+      "loss": 0.2633,
+      "step": 1216
+    },
+    {
+      "epoch": 1.6013157894736842,
+      "grad_norm": 0.07437930134164347,
+      "learning_rate": 0.00010566163609856117,
+      "loss": 0.2575,
+      "step": 1217
+    },
+    {
+      "epoch": 1.6026315789473684,
+      "grad_norm": 0.0780918664132209,
+      "learning_rate": 0.00010550877603558655,
+      "loss": 0.2647,
+      "step": 1218
+    },
+    {
+      "epoch": 1.6039473684210526,
+      "grad_norm": 0.07895904000053701,
+      "learning_rate": 0.00010535590306041732,
+      "loss": 0.2657,
+      "step": 1219
+    },
+    {
+      "epoch": 1.6052631578947367,
+      "grad_norm": 0.0782823182484623,
+      "learning_rate": 0.00010520301753137724,
+      "loss": 0.2539,
+      "step": 1220
+    },
+    {
+      "epoch": 1.606578947368421,
+      "grad_norm": 0.07836129489018223,
+      "learning_rate": 0.00010505011980681962,
+      "loss": 0.2623,
+      "step": 1221
+    },
+    {
+      "epoch": 1.6078947368421053,
+      "grad_norm": 0.07921294679317484,
+      "learning_rate": 0.00010489721024512618,
+      "loss": 0.2574,
+      "step": 1222
+    },
+    {
+      "epoch": 1.6092105263157894,
+      "grad_norm": 0.08253466147497233,
+      "learning_rate": 0.00010474428920470654,
+      "loss": 0.2626,
+      "step": 1223
+    },
+    {
+      "epoch": 1.6105263157894738,
+      "grad_norm": 0.0783379793278089,
+      "learning_rate": 0.00010459135704399718,
+      "loss": 0.2565,
+      "step": 1224
+    },
+    {
+      "epoch": 1.611842105263158,
+      "grad_norm": 0.0789556340245482,
+      "learning_rate": 0.00010443841412146065,
+      "loss": 0.2732,
+      "step": 1225
+    },
+    {
+      "epoch": 1.6131578947368421,
+      "grad_norm": 0.08284705629363337,
+      "learning_rate": 0.00010428546079558463,
+      "loss": 0.2784,
+      "step": 1226
+    },
+    {
+      "epoch": 1.6144736842105263,
+      "grad_norm": 0.07926378933409738,
+      "learning_rate": 0.00010413249742488131,
+      "loss": 0.2626,
+      "step": 1227
+    },
+    {
+      "epoch": 1.6157894736842104,
+      "grad_norm": 0.07646962136820679,
+      "learning_rate": 0.00010397952436788642,
+      "loss": 0.2576,
+      "step": 1228
+    },
+    {
+      "epoch": 1.6171052631578946,
+      "grad_norm": 0.07459202591852203,
+      "learning_rate": 0.00010382654198315834,
+      "loss": 0.2544,
+      "step": 1229
+    },
+    {
+      "epoch": 1.618421052631579,
+      "grad_norm": 0.07728933092230707,
+      "learning_rate": 0.00010367355062927726,
+      "loss": 0.2597,
+      "step": 1230
+    },
+    {
+      "epoch": 1.6197368421052631,
+      "grad_norm": 0.0740833624188333,
+      "learning_rate": 0.00010352055066484449,
+      "loss": 0.2415,
+      "step": 1231
+    },
+    {
+      "epoch": 1.6210526315789475,
+      "grad_norm": 0.07917560120385768,
+      "learning_rate": 0.00010336754244848157,
+      "loss": 0.2615,
+      "step": 1232
+    },
+    {
+      "epoch": 1.6223684210526317,
+      "grad_norm": 0.08063325789327866,
+      "learning_rate": 0.00010321452633882922,
+      "loss": 0.2741,
+      "step": 1233
+    },
+    {
+      "epoch": 1.6236842105263158,
+      "grad_norm": 0.07913725813766033,
+      "learning_rate": 0.00010306150269454675,
+      "loss": 0.2658,
+      "step": 1234
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 0.07936038810014337,
+      "learning_rate": 0.00010290847187431113,
+      "loss": 0.2618,
+      "step": 1235
+    },
+    {
+      "epoch": 1.6263157894736842,
+      "grad_norm": 0.07924213689691284,
+      "learning_rate": 0.00010275543423681621,
+      "loss": 0.2614,
+      "step": 1236
+    },
+    {
+      "epoch": 1.6276315789473683,
+      "grad_norm": 0.0771038728604427,
+      "learning_rate": 0.0001026023901407717,
+      "loss": 0.2477,
+      "step": 1237
+    },
+    {
+      "epoch": 1.6289473684210525,
+      "grad_norm": 0.0770735907526327,
+      "learning_rate": 0.00010244933994490249,
+      "loss": 0.2548,
+      "step": 1238
+    },
+    {
+      "epoch": 1.6302631578947369,
+      "grad_norm": 0.07674015928306312,
+      "learning_rate": 0.0001022962840079478,
+      "loss": 0.2547,
+      "step": 1239
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.07857136632734509,
+      "learning_rate": 0.00010214322268866032,
+      "loss": 0.2547,
+      "step": 1240
+    },
+    {
+      "epoch": 1.6328947368421054,
+      "grad_norm": 0.08094746714498062,
+      "learning_rate": 0.00010199015634580528,
+      "loss": 0.2612,
+      "step": 1241
+    },
+    {
+      "epoch": 1.6342105263157896,
+      "grad_norm": 0.07859907086130548,
+      "learning_rate": 0.00010183708533815974,
+      "loss": 0.2629,
+      "step": 1242
+    },
+    {
+      "epoch": 1.6355263157894737,
+      "grad_norm": 0.07847032246484141,
+      "learning_rate": 0.0001016840100245117,
+      "loss": 0.2587,
+      "step": 1243
+    },
+    {
+      "epoch": 1.6368421052631579,
+      "grad_norm": 0.0785820431028293,
+      "learning_rate": 0.00010153093076365923,
+      "loss": 0.2562,
+      "step": 1244
+    },
+    {
+      "epoch": 1.638157894736842,
+      "grad_norm": 0.07896059641278598,
+      "learning_rate": 0.00010137784791440965,
+      "loss": 0.2572,
+      "step": 1245
+    },
+    {
+      "epoch": 1.6394736842105262,
+      "grad_norm": 0.07921831824560291,
+      "learning_rate": 0.00010122476183557869,
+      "loss": 0.2727,
+      "step": 1246
+    },
+    {
+      "epoch": 1.6407894736842106,
+      "grad_norm": 0.07588261624391224,
+      "learning_rate": 0.00010107167288598967,
+      "loss": 0.2505,
+      "step": 1247
+    },
+    {
+      "epoch": 1.6421052631578947,
+      "grad_norm": 0.07634558208956514,
+      "learning_rate": 0.00010091858142447265,
+      "loss": 0.2536,
+      "step": 1248
+    },
+    {
+      "epoch": 1.643421052631579,
+      "grad_norm": 0.07689177759409037,
+      "learning_rate": 0.00010076548780986352,
+      "loss": 0.2603,
+      "step": 1249
+    },
+    {
+      "epoch": 1.6447368421052633,
+      "grad_norm": 0.07734671917849004,
+      "learning_rate": 0.00010061239240100327,
+      "loss": 0.2473,
+      "step": 1250
+    },
+    {
+      "epoch": 1.6460526315789474,
+      "grad_norm": 0.07734454423729424,
+      "learning_rate": 0.00010045929555673705,
+      "loss": 0.2687,
+      "step": 1251
+    },
+    {
+      "epoch": 1.6473684210526316,
+      "grad_norm": 0.0778519121851718,
+      "learning_rate": 0.00010030619763591347,
+      "loss": 0.2659,
+      "step": 1252
+    },
+    {
+      "epoch": 1.6486842105263158,
+      "grad_norm": 0.07985594655530279,
+      "learning_rate": 0.00010015309899738355,
+      "loss": 0.269,
+      "step": 1253
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.07695791648645144,
+      "learning_rate": 0.0001,
+      "loss": 0.2552,
+      "step": 1254
+    },
+    {
+      "epoch": 1.651315789473684,
+      "grad_norm": 0.07817327760226638,
+      "learning_rate": 9.984690100261648e-05,
+      "loss": 0.2493,
+      "step": 1255
+    },
+    {
+      "epoch": 1.6526315789473685,
+      "grad_norm": 0.07787762754684492,
+      "learning_rate": 9.969380236408656e-05,
+      "loss": 0.2553,
+      "step": 1256
+    },
+    {
+      "epoch": 1.6539473684210526,
+      "grad_norm": 0.08135305782817635,
+      "learning_rate": 9.954070444326293e-05,
+      "loss": 0.2505,
+      "step": 1257
+    },
+    {
+      "epoch": 1.655263157894737,
+      "grad_norm": 0.07789141129556089,
+      "learning_rate": 9.938760759899674e-05,
+      "loss": 0.2527,
+      "step": 1258
+    },
+    {
+      "epoch": 1.6565789473684212,
+      "grad_norm": 0.07824485146455672,
+      "learning_rate": 9.923451219013651e-05,
+      "loss": 0.2603,
+      "step": 1259
+    },
+    {
+      "epoch": 1.6578947368421053,
+      "grad_norm": 0.07858931038856287,
+      "learning_rate": 9.908141857552737e-05,
+      "loss": 0.2631,
+      "step": 1260
+    },
+    {
+      "epoch": 1.6592105263157895,
+      "grad_norm": 0.07644569879074736,
+      "learning_rate": 9.892832711401036e-05,
+      "loss": 0.2436,
+      "step": 1261
+    },
+    {
+      "epoch": 1.6605263157894736,
+      "grad_norm": 0.08015020052581098,
+      "learning_rate": 9.877523816442133e-05,
+      "loss": 0.2733,
+      "step": 1262
+    },
+    {
+      "epoch": 1.6618421052631578,
+      "grad_norm": 0.07973057338959542,
+      "learning_rate": 9.862215208559037e-05,
+      "loss": 0.2595,
+      "step": 1263
+    },
+    {
+      "epoch": 1.663157894736842,
+      "grad_norm": 0.07569295606836161,
+      "learning_rate": 9.846906923634079e-05,
+      "loss": 0.2425,
+      "step": 1264
+    },
+    {
+      "epoch": 1.6644736842105263,
+      "grad_norm": 0.07591010855329165,
+      "learning_rate": 9.831598997548831e-05,
+      "loss": 0.247,
+      "step": 1265
+    },
+    {
+      "epoch": 1.6657894736842105,
+      "grad_norm": 0.07750100103712705,
+      "learning_rate": 9.816291466184026e-05,
+      "loss": 0.2475,
+      "step": 1266
+    },
+    {
+      "epoch": 1.6671052631578949,
+      "grad_norm": 0.07706201585723965,
+      "learning_rate": 9.800984365419475e-05,
+      "loss": 0.2611,
+      "step": 1267
+    },
+    {
+      "epoch": 1.668421052631579,
+      "grad_norm": 0.07793486034207057,
+      "learning_rate": 9.78567773113397e-05,
+      "loss": 0.263,
+      "step": 1268
+    },
+    {
+      "epoch": 1.6697368421052632,
+      "grad_norm": 0.079099268943506,
+      "learning_rate": 9.770371599205222e-05,
+      "loss": 0.2578,
+      "step": 1269
+    },
+    {
+      "epoch": 1.6710526315789473,
+      "grad_norm": 0.07871797278447953,
+      "learning_rate": 9.755066005509753e-05,
+      "loss": 0.2578,
+      "step": 1270
+    },
+    {
+      "epoch": 1.6723684210526315,
+      "grad_norm": 0.0811458532081561,
+      "learning_rate": 9.739760985922832e-05,
+      "loss": 0.2579,
+      "step": 1271
+    },
+    {
+      "epoch": 1.6736842105263157,
+      "grad_norm": 0.07824873316859218,
+      "learning_rate": 9.724456576318381e-05,
+      "loss": 0.2754,
+      "step": 1272
+    },
+    {
+      "epoch": 1.675,
+      "grad_norm": 0.07647687339497086,
+      "learning_rate": 9.709152812568886e-05,
+      "loss": 0.2575,
+      "step": 1273
+    },
+    {
+      "epoch": 1.6763157894736842,
+      "grad_norm": 0.07842920421047767,
+      "learning_rate": 9.693849730545326e-05,
+      "loss": 0.248,
+      "step": 1274
+    },
+    {
+      "epoch": 1.6776315789473686,
+      "grad_norm": 0.0779966215397614,
+      "learning_rate": 9.678547366117083e-05,
+      "loss": 0.2669,
+      "step": 1275
+    },
+    {
+      "epoch": 1.6789473684210527,
+      "grad_norm": 0.07565197476267557,
+      "learning_rate": 9.663245755151846e-05,
+      "loss": 0.2553,
+      "step": 1276
+    },
+    {
+      "epoch": 1.680263157894737,
+      "grad_norm": 0.07727410664675037,
+      "learning_rate": 9.647944933515552e-05,
+      "loss": 0.2528,
+      "step": 1277
+    },
+    {
+      "epoch": 1.681578947368421,
+      "grad_norm": 0.076100890561838,
+      "learning_rate": 9.632644937072277e-05,
+      "loss": 0.2547,
+      "step": 1278
+    },
+    {
+      "epoch": 1.6828947368421052,
+      "grad_norm": 0.07998682208123038,
+      "learning_rate": 9.617345801684169e-05,
+      "loss": 0.2606,
+      "step": 1279
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.0766101279713053,
+      "learning_rate": 9.602047563211359e-05,
+      "loss": 0.2706,
+      "step": 1280
+    },
+    {
+      "epoch": 1.6855263157894735,
+      "grad_norm": 0.07714573829749818,
+      "learning_rate": 9.586750257511867e-05,
+      "loss": 0.2533,
+      "step": 1281
+    },
+    {
+      "epoch": 1.686842105263158,
+      "grad_norm": 0.0774718178209979,
+      "learning_rate": 9.571453920441538e-05,
+      "loss": 0.2541,
+      "step": 1282
+    },
+    {
+      "epoch": 1.688157894736842,
+      "grad_norm": 0.07724082139457383,
+      "learning_rate": 9.556158587853941e-05,
+      "loss": 0.2548,
+      "step": 1283
+    },
+    {
+      "epoch": 1.6894736842105265,
+      "grad_norm": 0.08138472870698339,
+      "learning_rate": 9.540864295600283e-05,
+      "loss": 0.2767,
+      "step": 1284
+    },
+    {
+      "epoch": 1.6907894736842106,
+      "grad_norm": 0.07668384972603022,
+      "learning_rate": 9.525571079529347e-05,
+      "loss": 0.2614,
+      "step": 1285
+    },
+    {
+      "epoch": 1.6921052631578948,
+      "grad_norm": 0.07631635135059317,
+      "learning_rate": 9.510278975487384e-05,
+      "loss": 0.2568,
+      "step": 1286
+    },
+    {
+      "epoch": 1.693421052631579,
+      "grad_norm": 0.07762293294175955,
+      "learning_rate": 9.49498801931804e-05,
+      "loss": 0.2608,
+      "step": 1287
+    },
+    {
+      "epoch": 1.694736842105263,
+      "grad_norm": 0.08091630947772432,
+      "learning_rate": 9.479698246862276e-05,
+      "loss": 0.2624,
+      "step": 1288
+    },
+    {
+      "epoch": 1.6960526315789473,
+      "grad_norm": 0.07844980001770621,
+      "learning_rate": 9.464409693958269e-05,
+      "loss": 0.2589,
+      "step": 1289
+    },
+    {
+      "epoch": 1.6973684210526314,
+      "grad_norm": 0.07835999657865879,
+      "learning_rate": 9.449122396441345e-05,
+      "loss": 0.2673,
+      "step": 1290
+    },
+    {
+      "epoch": 1.6986842105263158,
+      "grad_norm": 0.07660019194402308,
+      "learning_rate": 9.433836390143887e-05,
+      "loss": 0.2533,
+      "step": 1291
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.0775129862715817,
+      "learning_rate": 9.418551710895243e-05,
+      "loss": 0.2526,
+      "step": 1292
+    },
+    {
+      "epoch": 1.7013157894736843,
+      "grad_norm": 0.07852001442550643,
+      "learning_rate": 9.403268394521662e-05,
+      "loss": 0.2536,
+      "step": 1293
+    },
+    {
+      "epoch": 1.7026315789473685,
+      "grad_norm": 0.07775416745660803,
+      "learning_rate": 9.38798647684619e-05,
+      "loss": 0.2568,
+      "step": 1294
+    },
+    {
+      "epoch": 1.7039473684210527,
+      "grad_norm": 0.0773123834839056,
+      "learning_rate": 9.372705993688599e-05,
+      "loss": 0.2487,
+      "step": 1295
+    },
+    {
+      "epoch": 1.7052631578947368,
+      "grad_norm": 0.07933332672618959,
+      "learning_rate": 9.357426980865301e-05,
+      "loss": 0.2546,
+      "step": 1296
+    },
+    {
+      "epoch": 1.706578947368421,
+      "grad_norm": 0.0758849825907164,
+      "learning_rate": 9.342149474189251e-05,
+      "loss": 0.2527,
+      "step": 1297
+    },
+    {
+      "epoch": 1.7078947368421051,
+      "grad_norm": 0.07768939168993196,
+      "learning_rate": 9.326873509469887e-05,
+      "loss": 0.2635,
+      "step": 1298
+    },
+    {
+      "epoch": 1.7092105263157895,
+      "grad_norm": 0.07733832118219917,
+      "learning_rate": 9.311599122513029e-05,
+      "loss": 0.2639,
+      "step": 1299
+    },
+    {
+      "epoch": 1.7105263157894737,
+      "grad_norm": 0.07871573341153681,
+      "learning_rate": 9.296326349120785e-05,
+      "loss": 0.2618,
+      "step": 1300
+    },
+    {
+      "epoch": 1.711842105263158,
+      "grad_norm": 0.07866153909260622,
+      "learning_rate": 9.281055225091503e-05,
+      "loss": 0.2594,
+      "step": 1301
+    },
+    {
+      "epoch": 1.7131578947368422,
+      "grad_norm": 0.07974140917973452,
+      "learning_rate": 9.265785786219647e-05,
+      "loss": 0.2668,
+      "step": 1302
+    },
+    {
+      "epoch": 1.7144736842105264,
+      "grad_norm": 0.08096294253599645,
+      "learning_rate": 9.250518068295744e-05,
+      "loss": 0.2652,
+      "step": 1303
+    },
+    {
+      "epoch": 1.7157894736842105,
+      "grad_norm": 0.0790015551619781,
+      "learning_rate": 9.235252107106279e-05,
+      "loss": 0.2657,
+      "step": 1304
+    },
+    {
+      "epoch": 1.7171052631578947,
+      "grad_norm": 0.07763615541336244,
+      "learning_rate": 9.219987938433621e-05,
+      "loss": 0.2629,
+      "step": 1305
+    },
+    {
+      "epoch": 1.7184210526315788,
+      "grad_norm": 0.07778669559237536,
+      "learning_rate": 9.204725598055942e-05,
+      "loss": 0.2636,
+      "step": 1306
+    },
+    {
+      "epoch": 1.719736842105263,
+      "grad_norm": 0.07939255811714764,
+      "learning_rate": 9.189465121747125e-05,
+      "loss": 0.2705,
+      "step": 1307
+    },
+    {
+      "epoch": 1.7210526315789474,
+      "grad_norm": 0.07960488700289318,
+      "learning_rate": 9.174206545276677e-05,
+      "loss": 0.2706,
+      "step": 1308
+    },
+    {
+      "epoch": 1.7223684210526315,
+      "grad_norm": 0.07723537042473928,
+      "learning_rate": 9.158949904409668e-05,
+      "loss": 0.2621,
+      "step": 1309
+    },
+    {
+      "epoch": 1.723684210526316,
+      "grad_norm": 0.07861687285953221,
+      "learning_rate": 9.143695234906611e-05,
+      "loss": 0.2549,
+      "step": 1310
+    },
+    {
+      "epoch": 1.725,
+      "grad_norm": 0.07928259749395238,
+      "learning_rate": 9.128442572523417e-05,
+      "loss": 0.256,
+      "step": 1311
+    },
+    {
+      "epoch": 1.7263157894736842,
+      "grad_norm": 0.07977907458496993,
+      "learning_rate": 9.113191953011287e-05,
+      "loss": 0.2621,
+      "step": 1312
+    },
+    {
+      "epoch": 1.7276315789473684,
+      "grad_norm": 0.07908599126862541,
+      "learning_rate": 9.09794341211662e-05,
+      "loss": 0.2577,
+      "step": 1313
+    },
+    {
+      "epoch": 1.7289473684210526,
+      "grad_norm": 0.07804526868832733,
+      "learning_rate": 9.082696985580964e-05,
+      "loss": 0.2574,
+      "step": 1314
+    },
+    {
+      "epoch": 1.7302631578947367,
+      "grad_norm": 0.07671319689683427,
+      "learning_rate": 9.0674527091409e-05,
+      "loss": 0.2551,
+      "step": 1315
+    },
+    {
+      "epoch": 1.731578947368421,
+      "grad_norm": 0.07918675981206362,
+      "learning_rate": 9.052210618527966e-05,
+      "loss": 0.2478,
+      "step": 1316
+    },
+    {
+      "epoch": 1.7328947368421053,
+      "grad_norm": 0.07700717313469471,
+      "learning_rate": 9.036970749468584e-05,
+      "loss": 0.2484,
+      "step": 1317
+    },
+    {
+      "epoch": 1.7342105263157894,
+      "grad_norm": 0.07832287153888227,
+      "learning_rate": 9.021733137683962e-05,
+      "loss": 0.2604,
+      "step": 1318
+    },
+    {
+      "epoch": 1.7355263157894738,
+      "grad_norm": 0.076498834528713,
+      "learning_rate": 9.006497818890024e-05,
+      "loss": 0.2536,
+      "step": 1319
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.077411644809103,
+      "learning_rate": 8.991264828797319e-05,
+      "loss": 0.2578,
+      "step": 1320
+    },
+    {
+      "epoch": 1.7381578947368421,
+      "grad_norm": 0.07747513215833557,
+      "learning_rate": 8.97603420311093e-05,
+      "loss": 0.2574,
+      "step": 1321
+    },
+    {
+      "epoch": 1.7394736842105263,
+      "grad_norm": 0.07842803983473304,
+      "learning_rate": 8.960805977530404e-05,
+      "loss": 0.2606,
+      "step": 1322
+    },
+    {
+      "epoch": 1.7407894736842104,
+      "grad_norm": 0.07436646379437782,
+      "learning_rate": 8.945580187749666e-05,
+      "loss": 0.2487,
+      "step": 1323
+    },
+    {
+      "epoch": 1.7421052631578946,
+      "grad_norm": 0.07487978241239661,
+      "learning_rate": 8.930356869456919e-05,
+      "loss": 0.2532,
+      "step": 1324
+    },
+    {
+      "epoch": 1.743421052631579,
+      "grad_norm": 0.07966307306554148,
+      "learning_rate": 8.915136058334588e-05,
+      "loss": 0.2665,
+      "step": 1325
+    },
+    {
+      "epoch": 1.7447368421052631,
+      "grad_norm": 0.08106464265363358,
+      "learning_rate": 8.899917790059208e-05,
+      "loss": 0.2643,
+      "step": 1326
+    },
+    {
+      "epoch": 1.7460526315789475,
+      "grad_norm": 0.07593234358817623,
+      "learning_rate": 8.884702100301364e-05,
+      "loss": 0.2586,
+      "step": 1327
+    },
+    {
+      "epoch": 1.7473684210526317,
+      "grad_norm": 0.07609912132517359,
+      "learning_rate": 8.869489024725595e-05,
+      "loss": 0.2459,
+      "step": 1328
+    },
+    {
+      "epoch": 1.7486842105263158,
+      "grad_norm": 0.07772499392839312,
+      "learning_rate": 8.854278598990305e-05,
+      "loss": 0.2632,
+      "step": 1329
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.07646456091007017,
+      "learning_rate": 8.839070858747697e-05,
+      "loss": 0.2526,
+      "step": 1330
+    },
+    {
+      "epoch": 1.7513157894736842,
+      "grad_norm": 0.07793051011581553,
+      "learning_rate": 8.823865839643677e-05,
+      "loss": 0.2714,
+      "step": 1331
+    },
+    {
+      "epoch": 1.7526315789473683,
+      "grad_norm": 0.07790374700766114,
+      "learning_rate": 8.808663577317764e-05,
+      "loss": 0.2605,
+      "step": 1332
+    },
+    {
+      "epoch": 1.7539473684210525,
+      "grad_norm": 0.0800107435876805,
+      "learning_rate": 8.793464107403028e-05,
+      "loss": 0.2595,
+      "step": 1333
+    },
+    {
+      "epoch": 1.7552631578947369,
+      "grad_norm": 0.08006102878272488,
+      "learning_rate": 8.778267465525985e-05,
+      "loss": 0.2667,
+      "step": 1334
+    },
+    {
+      "epoch": 1.756578947368421,
+      "grad_norm": 0.07794460119809767,
+      "learning_rate": 8.763073687306524e-05,
+      "loss": 0.2506,
+      "step": 1335
+    },
+    {
+      "epoch": 1.7578947368421054,
+      "grad_norm": 0.07759219048166072,
+      "learning_rate": 8.747882808357828e-05,
+      "loss": 0.2705,
+      "step": 1336
+    },
+    {
+      "epoch": 1.7592105263157896,
+      "grad_norm": 0.07712939424044654,
+      "learning_rate": 8.732694864286273e-05,
+      "loss": 0.2635,
+      "step": 1337
+    },
+    {
+      "epoch": 1.7605263157894737,
+      "grad_norm": 0.07743375242887887,
+      "learning_rate": 8.717509890691368e-05,
+      "loss": 0.2587,
+      "step": 1338
+    },
+    {
+      "epoch": 1.7618421052631579,
+      "grad_norm": 0.07860695060081709,
+      "learning_rate": 8.702327923165654e-05,
+      "loss": 0.2661,
+      "step": 1339
+    },
+    {
+      "epoch": 1.763157894736842,
+      "grad_norm": 0.0792831934996456,
+      "learning_rate": 8.687148997294621e-05,
+      "loss": 0.2678,
+      "step": 1340
+    },
+    {
+      "epoch": 1.7644736842105262,
+      "grad_norm": 0.07976560061008468,
+      "learning_rate": 8.671973148656634e-05,
+      "loss": 0.2619,
+      "step": 1341
+    },
+    {
+      "epoch": 1.7657894736842106,
+      "grad_norm": 0.08106987042463963,
+      "learning_rate": 8.656800412822847e-05,
+      "loss": 0.2735,
+      "step": 1342
+    },
+    {
+      "epoch": 1.7671052631578947,
+      "grad_norm": 0.07818392747232804,
+      "learning_rate": 8.641630825357115e-05,
+      "loss": 0.2387,
+      "step": 1343
+    },
+    {
+      "epoch": 1.768421052631579,
+      "grad_norm": 0.07902258148724874,
+      "learning_rate": 8.626464421815919e-05,
+      "loss": 0.2524,
+      "step": 1344
+    },
+    {
+      "epoch": 1.7697368421052633,
+      "grad_norm": 0.07756812342250517,
+      "learning_rate": 8.611301237748267e-05,
+      "loss": 0.2621,
+      "step": 1345
+    },
+    {
+      "epoch": 1.7710526315789474,
+      "grad_norm": 0.07885482619098905,
+      "learning_rate": 8.596141308695628e-05,
+      "loss": 0.2544,
+      "step": 1346
+    },
+    {
+      "epoch": 1.7723684210526316,
+      "grad_norm": 0.08005110335934358,
+      "learning_rate": 8.580984670191848e-05,
+      "loss": 0.2539,
+      "step": 1347
+    },
+    {
+      "epoch": 1.7736842105263158,
+      "grad_norm": 0.07943085562482573,
+      "learning_rate": 8.565831357763039e-05,
+      "loss": 0.2622,
+      "step": 1348
+    },
+    {
+      "epoch": 1.775,
+      "grad_norm": 0.08121598066546899,
+      "learning_rate": 8.550681406927535e-05,
+      "loss": 0.2658,
+      "step": 1349
+    },
+    {
+      "epoch": 1.776315789473684,
+      "grad_norm": 0.0773683018126476,
+      "learning_rate": 8.535534853195786e-05,
+      "loss": 0.2629,
+      "step": 1350
+    },
+    {
+      "epoch": 1.7776315789473685,
+      "grad_norm": 0.07743762444071942,
+      "learning_rate": 8.520391732070279e-05,
+      "loss": 0.2477,
+      "step": 1351
+    },
+    {
+      "epoch": 1.7789473684210526,
+      "grad_norm": 0.07738788021837957,
+      "learning_rate": 8.505252079045458e-05,
+      "loss": 0.2617,
+      "step": 1352
+    },
+    {
+      "epoch": 1.780263157894737,
+      "grad_norm": 0.07732362511743367,
+      "learning_rate": 8.490115929607631e-05,
+      "loss": 0.2562,
+      "step": 1353
+    },
+    {
+      "epoch": 1.7815789473684212,
+      "grad_norm": 0.07882812579817991,
+      "learning_rate": 8.474983319234899e-05,
+      "loss": 0.26,
+      "step": 1354
+    },
+    {
+      "epoch": 1.7828947368421053,
+      "grad_norm": 0.0764856203909363,
+      "learning_rate": 8.459854283397073e-05,
+      "loss": 0.2561,
+      "step": 1355
+    },
+    {
+      "epoch": 1.7842105263157895,
+      "grad_norm": 0.07723245518951646,
+      "learning_rate": 8.444728857555572e-05,
+      "loss": 0.2504,
+      "step": 1356
+    },
+    {
+      "epoch": 1.7855263157894736,
+      "grad_norm": 0.07746876461441092,
+      "learning_rate": 8.42960707716336e-05,
+      "loss": 0.247,
+      "step": 1357
+    },
+    {
+      "epoch": 1.7868421052631578,
+      "grad_norm": 0.07569577111860483,
+      "learning_rate": 8.414488977664859e-05,
+      "loss": 0.2409,
+      "step": 1358
+    },
+    {
+      "epoch": 1.788157894736842,
+      "grad_norm": 0.07919043138051154,
+      "learning_rate": 8.399374594495861e-05,
+      "loss": 0.2483,
+      "step": 1359
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.08072885452301694,
+      "learning_rate": 8.384263963083453e-05,
+      "loss": 0.2527,
+      "step": 1360
+    },
+    {
+      "epoch": 1.7907894736842105,
+      "grad_norm": 0.07843210339508665,
+      "learning_rate": 8.369157118845914e-05,
+      "loss": 0.2538,
+      "step": 1361
+    },
+    {
+      "epoch": 1.7921052631578949,
+      "grad_norm": 0.08063967474828551,
+      "learning_rate": 8.35405409719266e-05,
+      "loss": 0.2619,
+      "step": 1362
+    },
+    {
+      "epoch": 1.793421052631579,
+      "grad_norm": 0.07796033453134701,
+      "learning_rate": 8.338954933524144e-05,
+      "loss": 0.253,
+      "step": 1363
+    },
+    {
+      "epoch": 1.7947368421052632,
+      "grad_norm": 0.08080704163282365,
+      "learning_rate": 8.323859663231768e-05,
+      "loss": 0.2691,
+      "step": 1364
+    },
+    {
+      "epoch": 1.7960526315789473,
+      "grad_norm": 0.0786006883745561,
+      "learning_rate": 8.308768321697815e-05,
+      "loss": 0.263,
+      "step": 1365
+    },
+    {
+      "epoch": 1.7973684210526315,
+      "grad_norm": 0.07705721459554137,
+      "learning_rate": 8.293680944295359e-05,
+      "loss": 0.2486,
+      "step": 1366
+    },
+    {
+      "epoch": 1.7986842105263157,
+      "grad_norm": 0.07751983519025603,
+      "learning_rate": 8.278597566388184e-05,
+      "loss": 0.2578,
+      "step": 1367
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.07892468926699212,
+      "learning_rate": 8.263518223330697e-05,
+      "loss": 0.2503,
+      "step": 1368
+    },
+    {
+      "epoch": 1.8013157894736842,
+      "grad_norm": 0.07633017824416555,
+      "learning_rate": 8.248442950467845e-05,
+      "loss": 0.2623,
+      "step": 1369
+    },
+    {
+      "epoch": 1.8026315789473686,
+      "grad_norm": 0.07718202591821682,
+      "learning_rate": 8.23337178313504e-05,
+      "loss": 0.2565,
+      "step": 1370
+    },
+    {
+      "epoch": 1.8039473684210527,
+      "grad_norm": 0.0807857739995025,
+      "learning_rate": 8.218304756658072e-05,
+      "loss": 0.2599,
+      "step": 1371
+    },
+    {
+      "epoch": 1.805263157894737,
+      "grad_norm": 0.0790429723121207,
+      "learning_rate": 8.203241906353014e-05,
+      "loss": 0.2577,
+      "step": 1372
+    },
+    {
+      "epoch": 1.806578947368421,
+      "grad_norm": 0.07931013296038247,
+      "learning_rate": 8.188183267526161e-05,
+      "loss": 0.2639,
+      "step": 1373
+    },
+    {
+      "epoch": 1.8078947368421052,
+      "grad_norm": 0.07877666664803827,
+      "learning_rate": 8.173128875473932e-05,
+      "loss": 0.2563,
+      "step": 1374
+    },
+    {
+      "epoch": 1.8092105263157894,
+      "grad_norm": 0.07782636186273496,
+      "learning_rate": 8.158078765482796e-05,
+      "loss": 0.2635,
+      "step": 1375
+    },
+    {
+      "epoch": 1.8105263157894735,
+      "grad_norm": 0.07632603628610292,
+      "learning_rate": 8.143032972829183e-05,
+      "loss": 0.2568,
+      "step": 1376
+    },
+    {
+      "epoch": 1.811842105263158,
+      "grad_norm": 0.07623514696597232,
+      "learning_rate": 8.127991532779401e-05,
+      "loss": 0.2498,
+      "step": 1377
+    },
+    {
+      "epoch": 1.813157894736842,
+      "grad_norm": 0.0777236902872697,
+      "learning_rate": 8.112954480589558e-05,
+      "loss": 0.2534,
+      "step": 1378
+    },
+    {
+      "epoch": 1.8144736842105265,
+      "grad_norm": 0.07656061477259994,
+      "learning_rate": 8.09792185150548e-05,
+      "loss": 0.26,
+      "step": 1379
+    },
+    {
+      "epoch": 1.8157894736842106,
+      "grad_norm": 0.0789133570152323,
+      "learning_rate": 8.082893680762619e-05,
+      "loss": 0.2604,
+      "step": 1380
+    },
+    {
+      "epoch": 1.8171052631578948,
+      "grad_norm": 0.07851474431282743,
+      "learning_rate": 8.067870003585978e-05,
+      "loss": 0.2663,
+      "step": 1381
+    },
+    {
+      "epoch": 1.818421052631579,
+      "grad_norm": 0.0789616006642964,
+      "learning_rate": 8.052850855190034e-05,
+      "loss": 0.2638,
+      "step": 1382
+    },
+    {
+      "epoch": 1.819736842105263,
+      "grad_norm": 0.07562051436427474,
+      "learning_rate": 8.037836270778642e-05,
+      "loss": 0.2512,
+      "step": 1383
+    },
+    {
+      "epoch": 1.8210526315789473,
+      "grad_norm": 0.08100206522865458,
+      "learning_rate": 8.022826285544968e-05,
+      "loss": 0.2474,
+      "step": 1384
+    },
+    {
+      "epoch": 1.8223684210526314,
+      "grad_norm": 0.0794818851301571,
+      "learning_rate": 8.007820934671383e-05,
+      "loss": 0.2713,
+      "step": 1385
+    },
+    {
+      "epoch": 1.8236842105263158,
+      "grad_norm": 0.07947535099199077,
+      "learning_rate": 7.992820253329409e-05,
+      "loss": 0.2589,
+      "step": 1386
+    },
+    {
+      "epoch": 1.825,
+      "grad_norm": 0.08012876631030338,
+      "learning_rate": 7.977824276679623e-05,
+      "loss": 0.2656,
+      "step": 1387
+    },
+    {
+      "epoch": 1.8263157894736843,
+      "grad_norm": 0.0778147716148084,
+      "learning_rate": 7.96283303987156e-05,
+      "loss": 0.2574,
+      "step": 1388
+    },
+    {
+      "epoch": 1.8276315789473685,
+      "grad_norm": 0.08108286483882464,
+      "learning_rate": 7.947846578043659e-05,
+      "loss": 0.2592,
+      "step": 1389
+    },
+    {
+      "epoch": 1.8289473684210527,
+      "grad_norm": 0.07898753250610759,
+      "learning_rate": 7.932864926323161e-05,
+      "loss": 0.2612,
+      "step": 1390
+    },
+    {
+      "epoch": 1.8302631578947368,
+      "grad_norm": 0.07750883606242706,
+      "learning_rate": 7.917888119826036e-05,
+      "loss": 0.2533,
+      "step": 1391
+    },
+    {
+      "epoch": 1.831578947368421,
+      "grad_norm": 0.08012157144092143,
+      "learning_rate": 7.902916193656898e-05,
+      "loss": 0.2664,
+      "step": 1392
+    },
+    {
+      "epoch": 1.8328947368421051,
+      "grad_norm": 0.07738904505017828,
+      "learning_rate": 7.887949182908912e-05,
+      "loss": 0.2599,
+      "step": 1393
+    },
+    {
+      "epoch": 1.8342105263157895,
+      "grad_norm": 0.07769928944177891,
+      "learning_rate": 7.872987122663733e-05,
+      "loss": 0.2624,
+      "step": 1394
+    },
+    {
+      "epoch": 1.8355263157894737,
+      "grad_norm": 0.08069668200882675,
+      "learning_rate": 7.858030047991411e-05,
+      "loss": 0.2589,
+      "step": 1395
+    },
+    {
+      "epoch": 1.836842105263158,
+      "grad_norm": 0.07833127833956029,
+      "learning_rate": 7.843077993950302e-05,
+      "loss": 0.2572,
+      "step": 1396
+    },
+    {
+      "epoch": 1.8381578947368422,
+      "grad_norm": 0.07971876020235151,
+      "learning_rate": 7.828130995586998e-05,
+      "loss": 0.2582,
+      "step": 1397
+    },
+    {
+      "epoch": 1.8394736842105264,
+      "grad_norm": 0.07885772994939849,
+      "learning_rate": 7.813189087936243e-05,
+      "loss": 0.2738,
+      "step": 1398
+    },
+    {
+      "epoch": 1.8407894736842105,
+      "grad_norm": 0.07736167226784646,
+      "learning_rate": 7.798252306020851e-05,
+      "loss": 0.2608,
+      "step": 1399
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.07630819815719928,
+      "learning_rate": 7.783320684851614e-05,
+      "loss": 0.2613,
+      "step": 1400
+    },
+    {
+      "epoch": 1.8434210526315788,
+      "grad_norm": 0.07834945368484907,
+      "learning_rate": 7.768394259427234e-05,
+      "loss": 0.2614,
+      "step": 1401
+    },
+    {
+      "epoch": 1.844736842105263,
+      "grad_norm": 0.07718022449050989,
+      "learning_rate": 7.753473064734232e-05,
+      "loss": 0.2704,
+      "step": 1402
+    },
+    {
+      "epoch": 1.8460526315789474,
+      "grad_norm": 0.07707441842552827,
+      "learning_rate": 7.738557135746873e-05,
+      "loss": 0.2576,
+      "step": 1403
+    },
+    {
+      "epoch": 1.8473684210526315,
+      "grad_norm": 0.07812383749123156,
+      "learning_rate": 7.72364650742707e-05,
+      "loss": 0.2635,
+      "step": 1404
+    },
+    {
+      "epoch": 1.848684210526316,
+      "grad_norm": 0.07972960299568377,
+      "learning_rate": 7.708741214724315e-05,
+      "loss": 0.2711,
+      "step": 1405
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.07567077698895516,
+      "learning_rate": 7.693841292575598e-05,
+      "loss": 0.2583,
+      "step": 1406
+    },
+    {
+      "epoch": 1.8513157894736842,
+      "grad_norm": 0.07920973523092241,
+      "learning_rate": 7.678946775905324e-05,
+      "loss": 0.2565,
+      "step": 1407
+    },
+    {
+      "epoch": 1.8526315789473684,
+      "grad_norm": 0.07722915516418472,
+      "learning_rate": 7.664057699625214e-05,
+      "loss": 0.2526,
+      "step": 1408
+    },
+    {
+      "epoch": 1.8539473684210526,
+      "grad_norm": 0.08254943723021463,
+      "learning_rate": 7.649174098634251e-05,
+      "loss": 0.27,
+      "step": 1409
+    },
+    {
+      "epoch": 1.8552631578947367,
+      "grad_norm": 0.0778109023166561,
+      "learning_rate": 7.634296007818576e-05,
+      "loss": 0.2516,
+      "step": 1410
+    },
+    {
+      "epoch": 1.856578947368421,
+      "grad_norm": 0.07700945488788534,
+      "learning_rate": 7.619423462051423e-05,
+      "loss": 0.2646,
+      "step": 1411
+    },
+    {
+      "epoch": 1.8578947368421053,
+      "grad_norm": 0.07630139704348335,
+      "learning_rate": 7.604556496193015e-05,
+      "loss": 0.2538,
+      "step": 1412
+    },
+    {
+      "epoch": 1.8592105263157894,
+      "grad_norm": 0.07939635050751274,
+      "learning_rate": 7.589695145090506e-05,
+      "loss": 0.2667,
+      "step": 1413
+    },
+    {
+      "epoch": 1.8605263157894738,
+      "grad_norm": 0.07616939180006002,
+      "learning_rate": 7.57483944357789e-05,
+      "loss": 0.2529,
+      "step": 1414
+    },
+    {
+      "epoch": 1.861842105263158,
+      "grad_norm": 0.08117405461306933,
+      "learning_rate": 7.559989426475917e-05,
+      "loss": 0.2555,
+      "step": 1415
+    },
+    {
+      "epoch": 1.8631578947368421,
+      "grad_norm": 0.08059557719232159,
+      "learning_rate": 7.54514512859201e-05,
+      "loss": 0.2628,
+      "step": 1416
+    },
+    {
+      "epoch": 1.8644736842105263,
+      "grad_norm": 0.078588193817039,
+      "learning_rate": 7.530306584720188e-05,
+      "loss": 0.2615,
+      "step": 1417
+    },
+    {
+      "epoch": 1.8657894736842104,
+      "grad_norm": 0.07865927809600427,
+      "learning_rate": 7.515473829640987e-05,
+      "loss": 0.2588,
+      "step": 1418
+    },
+    {
+      "epoch": 1.8671052631578946,
+      "grad_norm": 0.07875855234210856,
+      "learning_rate": 7.500646898121373e-05,
+      "loss": 0.254,
+      "step": 1419
+    },
+    {
+      "epoch": 1.868421052631579,
+      "grad_norm": 0.07994170191297952,
+      "learning_rate": 7.485825824914659e-05,
+      "loss": 0.2625,
+      "step": 1420
+    },
+    {
+      "epoch": 1.8697368421052631,
+      "grad_norm": 0.07669419084742685,
+      "learning_rate": 7.471010644760421e-05,
+      "loss": 0.2639,
+      "step": 1421
+    },
+    {
+      "epoch": 1.8710526315789475,
+      "grad_norm": 0.07671532012322799,
+      "learning_rate": 7.456201392384436e-05,
+      "loss": 0.2618,
+      "step": 1422
+    },
+    {
+      "epoch": 1.8723684210526317,
+      "grad_norm": 0.0760382420175465,
+      "learning_rate": 7.441398102498582e-05,
+      "loss": 0.2586,
+      "step": 1423
+    },
+    {
+      "epoch": 1.8736842105263158,
+      "grad_norm": 0.07479447004096319,
+      "learning_rate": 7.426600809800752e-05,
+      "loss": 0.2487,
+      "step": 1424
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.07952054209905557,
+      "learning_rate": 7.411809548974792e-05,
+      "loss": 0.2694,
+      "step": 1425
+    },
+    {
+      "epoch": 1.8763157894736842,
+      "grad_norm": 0.0795536063574035,
+      "learning_rate": 7.397024354690408e-05,
+      "loss": 0.2575,
+      "step": 1426
+    },
+    {
+      "epoch": 1.8776315789473683,
+      "grad_norm": 0.07893539617767596,
+      "learning_rate": 7.382245261603088e-05,
+      "loss": 0.271,
+      "step": 1427
+    },
+    {
+      "epoch": 1.8789473684210525,
+      "grad_norm": 0.08003786520908833,
+      "learning_rate": 7.36747230435401e-05,
+      "loss": 0.2499,
+      "step": 1428
+    },
+    {
+      "epoch": 1.8802631578947369,
+      "grad_norm": 0.078791451136056,
+      "learning_rate": 7.352705517569977e-05,
+      "loss": 0.2674,
+      "step": 1429
+    },
+    {
+      "epoch": 1.881578947368421,
+      "grad_norm": 0.0761073189754986,
+      "learning_rate": 7.337944935863333e-05,
+      "loss": 0.2561,
+      "step": 1430
+    },
+    {
+      "epoch": 1.8828947368421054,
+      "grad_norm": 0.07783417880514816,
+      "learning_rate": 7.32319059383187e-05,
+      "loss": 0.2566,
+      "step": 1431
+    },
+    {
+      "epoch": 1.8842105263157896,
+      "grad_norm": 0.07782790823841897,
+      "learning_rate": 7.308442526058756e-05,
+      "loss": 0.2681,
+      "step": 1432
+    },
+    {
+      "epoch": 1.8855263157894737,
+      "grad_norm": 0.07722518403441637,
+      "learning_rate": 7.293700767112458e-05,
+      "loss": 0.2581,
+      "step": 1433
+    },
+    {
+      "epoch": 1.8868421052631579,
+      "grad_norm": 0.07897906723081582,
+      "learning_rate": 7.278965351546648e-05,
+      "loss": 0.2618,
+      "step": 1434
+    },
+    {
+      "epoch": 1.888157894736842,
+      "grad_norm": 0.07503201774332066,
+      "learning_rate": 7.264236313900141e-05,
+      "loss": 0.2449,
+      "step": 1435
+    },
+    {
+      "epoch": 1.8894736842105262,
+      "grad_norm": 0.07528777047208206,
+      "learning_rate": 7.249513688696786e-05,
+      "loss": 0.2509,
+      "step": 1436
+    },
+    {
+      "epoch": 1.8907894736842106,
+      "grad_norm": 0.07701172257337681,
+      "learning_rate": 7.234797510445411e-05,
+      "loss": 0.2502,
+      "step": 1437
+    },
+    {
+      "epoch": 1.8921052631578947,
+      "grad_norm": 0.07610174570738121,
+      "learning_rate": 7.220087813639736e-05,
+      "loss": 0.2551,
+      "step": 1438
+    },
+    {
+      "epoch": 1.893421052631579,
+      "grad_norm": 0.07662368675063383,
+      "learning_rate": 7.205384632758285e-05,
+      "loss": 0.2611,
+      "step": 1439
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.07837947549366885,
+      "learning_rate": 7.190688002264308e-05,
+      "loss": 0.2587,
+      "step": 1440
+    },
+    {
+      "epoch": 1.8960526315789474,
+      "grad_norm": 0.07828653663313988,
+      "learning_rate": 7.175997956605701e-05,
+      "loss": 0.2542,
+      "step": 1441
+    },
+    {
+      "epoch": 1.8973684210526316,
+      "grad_norm": 0.07749211808620897,
+      "learning_rate": 7.161314530214931e-05,
+      "loss": 0.2554,
+      "step": 1442
+    },
+    {
+      "epoch": 1.8986842105263158,
+      "grad_norm": 0.07955661477515108,
+      "learning_rate": 7.146637757508949e-05,
+      "loss": 0.2465,
+      "step": 1443
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.0780534289115777,
+      "learning_rate": 7.131967672889101e-05,
+      "loss": 0.2587,
+      "step": 1444
+    },
+    {
+      "epoch": 1.901315789473684,
+      "grad_norm": 0.08230240962537452,
+      "learning_rate": 7.117304310741062e-05,
+      "loss": 0.2676,
+      "step": 1445
+    },
+    {
+      "epoch": 1.9026315789473685,
+      "grad_norm": 0.07514699270244664,
+      "learning_rate": 7.102647705434756e-05,
+      "loss": 0.2515,
+      "step": 1446
+    },
+    {
+      "epoch": 1.9039473684210526,
+      "grad_norm": 0.07838555881596923,
+      "learning_rate": 7.087997891324262e-05,
+      "loss": 0.2676,
+      "step": 1447
+    },
+    {
+      "epoch": 1.905263157894737,
+      "grad_norm": 0.07794040393822359,
+      "learning_rate": 7.073354902747741e-05,
+      "loss": 0.2622,
+      "step": 1448
+    },
+    {
+      "epoch": 1.9065789473684212,
+      "grad_norm": 0.07793028180395223,
+      "learning_rate": 7.058718774027364e-05,
+      "loss": 0.2581,
+      "step": 1449
+    },
+    {
+      "epoch": 1.9078947368421053,
+      "grad_norm": 0.07468275561587644,
+      "learning_rate": 7.044089539469212e-05,
+      "loss": 0.2536,
+      "step": 1450
+    },
+    {
+      "epoch": 1.9092105263157895,
+      "grad_norm": 0.07736426080010585,
+      "learning_rate": 7.029467233363216e-05,
+      "loss": 0.2607,
+      "step": 1451
+    },
+    {
+      "epoch": 1.9105263157894736,
+      "grad_norm": 0.07788040509177954,
+      "learning_rate": 7.014851889983057e-05,
+      "loss": 0.2699,
+      "step": 1452
+    },
+    {
+      "epoch": 1.9118421052631578,
+      "grad_norm": 0.07482713534215896,
+      "learning_rate": 7.000243543586102e-05,
+      "loss": 0.2562,
+      "step": 1453
+    },
+    {
+      "epoch": 1.913157894736842,
+      "grad_norm": 0.07738656143600638,
+      "learning_rate": 6.985642228413316e-05,
+      "loss": 0.2589,
+      "step": 1454
+    },
+    {
+      "epoch": 1.9144736842105263,
+      "grad_norm": 0.0771729364408806,
+      "learning_rate": 6.971047978689189e-05,
+      "loss": 0.2651,
+      "step": 1455
+    },
+    {
+      "epoch": 1.9157894736842105,
+      "grad_norm": 0.07648368441271167,
+      "learning_rate": 6.95646082862164e-05,
+      "loss": 0.2568,
+      "step": 1456
+    },
+    {
+      "epoch": 1.9171052631578949,
+      "grad_norm": 0.07769870758338829,
+      "learning_rate": 6.941880812401956e-05,
+      "loss": 0.2537,
+      "step": 1457
+    },
+    {
+      "epoch": 1.918421052631579,
+      "grad_norm": 0.07903999465640997,
+      "learning_rate": 6.927307964204694e-05,
+      "loss": 0.2666,
+      "step": 1458
+    },
+    {
+      "epoch": 1.9197368421052632,
+      "grad_norm": 0.08045066676431731,
+      "learning_rate": 6.912742318187624e-05,
+      "loss": 0.2605,
+      "step": 1459
+    },
+    {
+      "epoch": 1.9210526315789473,
+      "grad_norm": 0.07627530893025287,
+      "learning_rate": 6.898183908491617e-05,
+      "loss": 0.255,
+      "step": 1460
+    },
+    {
+      "epoch": 1.9223684210526315,
+      "grad_norm": 0.07918035796815091,
+      "learning_rate": 6.883632769240589e-05,
+      "loss": 0.2565,
+      "step": 1461
+    },
+    {
+      "epoch": 1.9236842105263157,
+      "grad_norm": 0.079331461000813,
+      "learning_rate": 6.869088934541419e-05,
+      "loss": 0.2595,
+      "step": 1462
+    },
+    {
+      "epoch": 1.925,
+      "grad_norm": 0.07819193010414092,
+      "learning_rate": 6.854552438483865e-05,
+      "loss": 0.2602,
+      "step": 1463
+    },
+    {
+      "epoch": 1.9263157894736842,
+      "grad_norm": 0.08127166962507164,
+      "learning_rate": 6.840023315140475e-05,
+      "loss": 0.2563,
+      "step": 1464
+    },
+    {
+      "epoch": 1.9276315789473686,
+      "grad_norm": 0.08070961078490956,
+      "learning_rate": 6.825501598566525e-05,
+      "loss": 0.259,
+      "step": 1465
+    },
+    {
+      "epoch": 1.9289473684210527,
+      "grad_norm": 0.0766943472529489,
+      "learning_rate": 6.810987322799926e-05,
+      "loss": 0.2586,
+      "step": 1466
+    },
+    {
+      "epoch": 1.930263157894737,
+      "grad_norm": 0.08113329672328959,
+      "learning_rate": 6.79648052186115e-05,
+      "loss": 0.2425,
+      "step": 1467
+    },
+    {
+      "epoch": 1.931578947368421,
+      "grad_norm": 0.07603940923660328,
+      "learning_rate": 6.781981229753145e-05,
+      "loss": 0.2534,
+      "step": 1468
+    },
+    {
+      "epoch": 1.9328947368421052,
+      "grad_norm": 0.07646855224356787,
+      "learning_rate": 6.76748948046126e-05,
+      "loss": 0.2506,
+      "step": 1469
+    },
+    {
+      "epoch": 1.9342105263157894,
+      "grad_norm": 0.07477033930912226,
+      "learning_rate": 6.753005307953167e-05,
+      "loss": 0.2509,
+      "step": 1470
+    },
+    {
+      "epoch": 1.9355263157894735,
+      "grad_norm": 0.07717545781453299,
+      "learning_rate": 6.738528746178776e-05,
+      "loss": 0.2576,
+      "step": 1471
+    },
+    {
+      "epoch": 1.936842105263158,
+      "grad_norm": 0.07742807164222197,
+      "learning_rate": 6.724059829070158e-05,
+      "loss": 0.2647,
+      "step": 1472
+    },
+    {
+      "epoch": 1.938157894736842,
+      "grad_norm": 0.07798070076914343,
+      "learning_rate": 6.709598590541469e-05,
+      "loss": 0.2593,
+      "step": 1473
+    },
+    {
+      "epoch": 1.9394736842105265,
+      "grad_norm": 0.07717711873805973,
+      "learning_rate": 6.695145064488861e-05,
+      "loss": 0.2589,
+      "step": 1474
+    },
+    {
+      "epoch": 1.9407894736842106,
+      "grad_norm": 0.07645570062299177,
+      "learning_rate": 6.680699284790415e-05,
+      "loss": 0.2612,
+      "step": 1475
+    },
+    {
+      "epoch": 1.9421052631578948,
+      "grad_norm": 0.07709702657938383,
+      "learning_rate": 6.666261285306047e-05,
+      "loss": 0.2643,
+      "step": 1476
+    },
+    {
+      "epoch": 1.943421052631579,
+      "grad_norm": 0.07634109328627783,
+      "learning_rate": 6.651831099877444e-05,
+      "loss": 0.2468,
+      "step": 1477
+    },
+    {
+      "epoch": 1.944736842105263,
+      "grad_norm": 0.07783850878336081,
+      "learning_rate": 6.637408762327972e-05,
+      "loss": 0.2607,
+      "step": 1478
+    },
+    {
+      "epoch": 1.9460526315789473,
+      "grad_norm": 0.07820503016401602,
+      "learning_rate": 6.622994306462611e-05,
+      "loss": 0.2574,
+      "step": 1479
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.0796202316851204,
+      "learning_rate": 6.608587766067852e-05,
+      "loss": 0.2575,
+      "step": 1480
+    },
+    {
+      "epoch": 1.9486842105263158,
+      "grad_norm": 0.07605261401674518,
+      "learning_rate": 6.59418917491165e-05,
+      "loss": 0.2591,
+      "step": 1481
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.07560659095541633,
+      "learning_rate": 6.579798566743314e-05,
+      "loss": 0.2526,
+      "step": 1482
+    },
+    {
+      "epoch": 1.9513157894736843,
+      "grad_norm": 0.07747930757786237,
+      "learning_rate": 6.565415975293448e-05,
+      "loss": 0.2631,
+      "step": 1483
+    },
+    {
+      "epoch": 1.9526315789473685,
+      "grad_norm": 0.07643184520906729,
+      "learning_rate": 6.551041434273861e-05,
+      "loss": 0.2508,
+      "step": 1484
+    },
+    {
+      "epoch": 1.9539473684210527,
+      "grad_norm": 0.0796433446523202,
+      "learning_rate": 6.536674977377496e-05,
+      "loss": 0.2543,
+      "step": 1485
+    },
+    {
+      "epoch": 1.9552631578947368,
+      "grad_norm": 0.07657952724716138,
+      "learning_rate": 6.522316638278347e-05,
+      "loss": 0.2552,
+      "step": 1486
+    },
+    {
+      "epoch": 1.956578947368421,
+      "grad_norm": 0.07677380083558996,
+      "learning_rate": 6.507966450631382e-05,
+      "loss": 0.2541,
+      "step": 1487
+    },
+    {
+      "epoch": 1.9578947368421051,
+      "grad_norm": 0.07595336702259202,
+      "learning_rate": 6.493624448072457e-05,
+      "loss": 0.2487,
+      "step": 1488
+    },
+    {
+      "epoch": 1.9592105263157895,
+      "grad_norm": 0.07721242087118127,
+      "learning_rate": 6.479290664218247e-05,
+      "loss": 0.2507,
+      "step": 1489
+    },
+    {
+      "epoch": 1.9605263157894737,
+      "grad_norm": 0.07755079832687756,
+      "learning_rate": 6.464965132666163e-05,
+      "loss": 0.2574,
+      "step": 1490
+    },
+    {
+      "epoch": 1.961842105263158,
+      "grad_norm": 0.07980082932294216,
+      "learning_rate": 6.450647886994272e-05,
+      "loss": 0.2704,
+      "step": 1491
+    },
+    {
+      "epoch": 1.9631578947368422,
+      "grad_norm": 0.07840280028957075,
+      "learning_rate": 6.43633896076122e-05,
+      "loss": 0.2565,
+      "step": 1492
+    },
+    {
+      "epoch": 1.9644736842105264,
+      "grad_norm": 0.07780809182938828,
+      "learning_rate": 6.422038387506149e-05,
+      "loss": 0.2626,
+      "step": 1493
+    },
+    {
+      "epoch": 1.9657894736842105,
+      "grad_norm": 0.07710732052984494,
+      "learning_rate": 6.407746200748628e-05,
+      "loss": 0.2565,
+      "step": 1494
+    },
+    {
+      "epoch": 1.9671052631578947,
+      "grad_norm": 0.07590458797290023,
+      "learning_rate": 6.393462433988569e-05,
+      "loss": 0.2547,
+      "step": 1495
+    },
+    {
+      "epoch": 1.9684210526315788,
+      "grad_norm": 0.07792407917489579,
+      "learning_rate": 6.379187120706138e-05,
+      "loss": 0.2466,
+      "step": 1496
+    },
+    {
+      "epoch": 1.969736842105263,
+      "grad_norm": 0.07781793554961416,
+      "learning_rate": 6.3649202943617e-05,
+      "loss": 0.266,
+      "step": 1497
+    },
+    {
+      "epoch": 1.9710526315789474,
+      "grad_norm": 0.07455889127205702,
+      "learning_rate": 6.350661988395723e-05,
+      "loss": 0.2458,
+      "step": 1498
+    },
+    {
+      "epoch": 1.9723684210526315,
+      "grad_norm": 0.07493719255336764,
+      "learning_rate": 6.336412236228697e-05,
+      "loss": 0.2565,
+      "step": 1499
+    },
+    {
+      "epoch": 1.973684210526316,
+      "grad_norm": 0.07690525334306028,
+      "learning_rate": 6.322171071261071e-05,
+      "loss": 0.2613,
+      "step": 1500
+    },
+    {
+      "epoch": 1.975,
+      "grad_norm": 0.07579197410852237,
+      "learning_rate": 6.307938526873157e-05,
+      "loss": 0.2523,
+      "step": 1501
+    },
+    {
+      "epoch": 1.9763157894736842,
+      "grad_norm": 0.08008252829550877,
+      "learning_rate": 6.293714636425071e-05,
+      "loss": 0.2671,
+      "step": 1502
+    },
+    {
+      "epoch": 1.9776315789473684,
+      "grad_norm": 0.07781399370339455,
+      "learning_rate": 6.279499433256642e-05,
+      "loss": 0.2603,
+      "step": 1503
+    },
+    {
+      "epoch": 1.9789473684210526,
+      "grad_norm": 0.07813006074806125,
+      "learning_rate": 6.26529295068733e-05,
+      "loss": 0.2493,
+      "step": 1504
+    },
+    {
+      "epoch": 1.9802631578947367,
+      "grad_norm": 0.07894579521693036,
+      "learning_rate": 6.251095222016162e-05,
+      "loss": 0.2624,
+      "step": 1505
+    },
+    {
+      "epoch": 1.981578947368421,
+      "grad_norm": 0.07494640550743084,
+      "learning_rate": 6.236906280521646e-05,
+      "loss": 0.2409,
+      "step": 1506
+    },
+    {
+      "epoch": 1.9828947368421053,
+      "grad_norm": 0.07893332158347492,
+      "learning_rate": 6.22272615946169e-05,
+      "loss": 0.2629,
+      "step": 1507
+    },
+    {
+      "epoch": 1.9842105263157894,
+      "grad_norm": 0.07842421722733443,
+      "learning_rate": 6.208554892073528e-05,
+      "loss": 0.2504,
+      "step": 1508
+    },
+    {
+      "epoch": 1.9855263157894738,
+      "grad_norm": 0.08117542344228214,
+      "learning_rate": 6.19439251157364e-05,
+      "loss": 0.2653,
+      "step": 1509
+    },
+    {
+      "epoch": 1.986842105263158,
+      "grad_norm": 0.07755918045728276,
+      "learning_rate": 6.180239051157681e-05,
+      "loss": 0.2558,
+      "step": 1510
+    },
+    {
+      "epoch": 1.9881578947368421,
+      "grad_norm": 0.07645794314850728,
+      "learning_rate": 6.166094544000398e-05,
+      "loss": 0.2559,
+      "step": 1511
+    },
+    {
+      "epoch": 1.9894736842105263,
+      "grad_norm": 0.07640733998010549,
+      "learning_rate": 6.151959023255545e-05,
+      "loss": 0.2537,
+      "step": 1512
+    },
+    {
+      "epoch": 1.9907894736842104,
+      "grad_norm": 0.0768529160978783,
+      "learning_rate": 6.137832522055817e-05,
+      "loss": 0.2601,
+      "step": 1513
+    },
+    {
+      "epoch": 1.9921052631578946,
+      "grad_norm": 0.07847739643759478,
+      "learning_rate": 6.123715073512772e-05,
+      "loss": 0.255,
+      "step": 1514
+    },
+    {
+      "epoch": 1.993421052631579,
+      "grad_norm": 0.07449855656212337,
+      "learning_rate": 6.109606710716741e-05,
+      "loss": 0.239,
+      "step": 1515
+    },
+    {
+      "epoch": 1.9947368421052631,
+      "grad_norm": 0.07846660941255253,
+      "learning_rate": 6.095507466736763e-05,
+      "loss": 0.2551,
+      "step": 1516
+    },
+    {
+      "epoch": 1.9960526315789475,
+      "grad_norm": 0.07789468698938513,
+      "learning_rate": 6.0814173746205e-05,
+      "loss": 0.2694,
+      "step": 1517
+    },
+    {
+      "epoch": 1.9973684210526317,
+      "grad_norm": 0.077200397264026,
+      "learning_rate": 6.067336467394169e-05,
+      "loss": 0.2589,
+      "step": 1518
+    },
+    {
+      "epoch": 1.9986842105263158,
+      "grad_norm": 0.08246944198719414,
+      "learning_rate": 6.0532647780624554e-05,
+      "loss": 0.2656,
+      "step": 1519
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.07831394886097501,
+      "learning_rate": 6.039202339608432e-05,
+      "loss": 0.2656,
+      "step": 1520
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.2646636962890625,
+      "eval_runtime": 136.4654,
+      "eval_samples_per_second": 37.504,
+      "eval_steps_per_second": 1.172,
+      "step": 1520
+    },
+    {
+      "epoch": 2.001315789473684,
+      "grad_norm": 0.07478380723960856,
+      "learning_rate": 6.025149184993498e-05,
+      "loss": 0.2524,
+      "step": 1521
+    },
+    {
+      "epoch": 2.0026315789473683,
+      "grad_norm": 0.07372772894034235,
+      "learning_rate": 6.011105347157289e-05,
+      "loss": 0.2399,
+      "step": 1522
+    },
+    {
+      "epoch": 2.0039473684210525,
+      "grad_norm": 0.07430175019575523,
+      "learning_rate": 5.9970708590175986e-05,
+      "loss": 0.2369,
+      "step": 1523
+    },
+    {
+      "epoch": 2.0052631578947366,
+      "grad_norm": 0.07626311971922609,
+      "learning_rate": 5.983045753470308e-05,
+      "loss": 0.2371,
+      "step": 1524
+    },
+    {
+      "epoch": 2.0065789473684212,
+      "grad_norm": 0.07792913324419348,
+      "learning_rate": 5.969030063389305e-05,
+      "loss": 0.2372,
+      "step": 1525
+    },
+    {
+      "epoch": 2.0078947368421054,
+      "grad_norm": 0.07819835291288194,
+      "learning_rate": 5.955023821626411e-05,
+      "loss": 0.2299,
+      "step": 1526
+    },
+    {
+      "epoch": 2.0092105263157896,
+      "grad_norm": 0.0782044620700486,
+      "learning_rate": 5.941027061011303e-05,
+      "loss": 0.243,
+      "step": 1527
+    },
+    {
+      "epoch": 2.0105263157894737,
+      "grad_norm": 0.08211154278042845,
+      "learning_rate": 5.927039814351426e-05,
+      "loss": 0.2432,
+      "step": 1528
+    },
+    {
+      "epoch": 2.011842105263158,
+      "grad_norm": 0.08053029149065172,
+      "learning_rate": 5.9130621144319334e-05,
+      "loss": 0.2349,
+      "step": 1529
+    },
+    {
+      "epoch": 2.013157894736842,
+      "grad_norm": 0.08170774492319388,
+      "learning_rate": 5.8990939940156e-05,
+      "loss": 0.2455,
+      "step": 1530
+    },
+    {
+      "epoch": 2.014473684210526,
+      "grad_norm": 0.07949305041445266,
+      "learning_rate": 5.885135485842743e-05,
+      "loss": 0.2351,
+      "step": 1531
+    },
+    {
+      "epoch": 2.0157894736842104,
+      "grad_norm": 0.08110005829531224,
+      "learning_rate": 5.8711866226311553e-05,
+      "loss": 0.2579,
+      "step": 1532
+    },
+    {
+      "epoch": 2.017105263157895,
+      "grad_norm": 0.08333787912458002,
+      "learning_rate": 5.857247437076012e-05,
+      "loss": 0.2387,
+      "step": 1533
+    },
+    {
+      "epoch": 2.018421052631579,
+      "grad_norm": 0.08010741675929599,
+      "learning_rate": 5.843317961849818e-05,
+      "loss": 0.2299,
+      "step": 1534
+    },
+    {
+      "epoch": 2.0197368421052633,
+      "grad_norm": 0.08205136086453177,
+      "learning_rate": 5.829398229602312e-05,
+      "loss": 0.2422,
+      "step": 1535
+    },
+    {
+      "epoch": 2.0210526315789474,
+      "grad_norm": 0.08134779329404473,
+      "learning_rate": 5.8154882729603876e-05,
+      "loss": 0.2488,
+      "step": 1536
+    },
+    {
+      "epoch": 2.0223684210526316,
+      "grad_norm": 0.0793049391780135,
+      "learning_rate": 5.8015881245280436e-05,
+      "loss": 0.2404,
+      "step": 1537
+    },
+    {
+      "epoch": 2.0236842105263158,
+      "grad_norm": 0.07994738843779876,
+      "learning_rate": 5.787697816886273e-05,
+      "loss": 0.2309,
+      "step": 1538
+    },
+    {
+      "epoch": 2.025,
+      "grad_norm": 0.07995832117150292,
+      "learning_rate": 5.773817382593008e-05,
+      "loss": 0.2243,
+      "step": 1539
+    },
+    {
+      "epoch": 2.026315789473684,
+      "grad_norm": 0.0805658044178454,
+      "learning_rate": 5.7599468541830356e-05,
+      "loss": 0.236,
+      "step": 1540
+    },
+    {
+      "epoch": 2.0276315789473682,
+      "grad_norm": 0.082233376699682,
+      "learning_rate": 5.7460862641679316e-05,
+      "loss": 0.2395,
+      "step": 1541
+    },
+    {
+      "epoch": 2.028947368421053,
+      "grad_norm": 0.08455600799602911,
+      "learning_rate": 5.732235645035964e-05,
+      "loss": 0.2527,
+      "step": 1542
+    },
+    {
+      "epoch": 2.030263157894737,
+      "grad_norm": 0.08166867758119097,
+      "learning_rate": 5.7183950292520473e-05,
+      "loss": 0.2247,
+      "step": 1543
+    },
+    {
+      "epoch": 2.031578947368421,
+      "grad_norm": 0.08362795390398718,
+      "learning_rate": 5.7045644492576346e-05,
+      "loss": 0.2334,
+      "step": 1544
+    },
+    {
+      "epoch": 2.0328947368421053,
+      "grad_norm": 0.08759943221985574,
+      "learning_rate": 5.690743937470657e-05,
+      "loss": 0.25,
+      "step": 1545
+    },
+    {
+      "epoch": 2.0342105263157895,
+      "grad_norm": 0.0834958876411245,
+      "learning_rate": 5.676933526285457e-05,
+      "loss": 0.2416,
+      "step": 1546
+    },
+    {
+      "epoch": 2.0355263157894736,
+      "grad_norm": 0.08052792625700166,
+      "learning_rate": 5.663133248072692e-05,
+      "loss": 0.2363,
+      "step": 1547
+    },
+    {
+      "epoch": 2.036842105263158,
+      "grad_norm": 0.08077196722556565,
+      "learning_rate": 5.64934313517927e-05,
+      "loss": 0.2395,
+      "step": 1548
+    },
+    {
+      "epoch": 2.038157894736842,
+      "grad_norm": 0.08231061675764222,
+      "learning_rate": 5.635563219928275e-05,
+      "loss": 0.2467,
+      "step": 1549
+    },
+    {
+      "epoch": 2.039473684210526,
+      "grad_norm": 0.08140686496008916,
+      "learning_rate": 5.62179353461888e-05,
+      "loss": 0.2399,
+      "step": 1550
+    },
+    {
+      "epoch": 2.0407894736842107,
+      "grad_norm": 0.08075751241578549,
+      "learning_rate": 5.608034111526298e-05,
+      "loss": 0.2262,
+      "step": 1551
+    },
+    {
+      "epoch": 2.042105263157895,
+      "grad_norm": 0.0814531147960288,
+      "learning_rate": 5.5942849829016695e-05,
+      "loss": 0.2292,
+      "step": 1552
+    },
+    {
+      "epoch": 2.043421052631579,
+      "grad_norm": 0.0807479054094009,
+      "learning_rate": 5.580546180972011e-05,
+      "loss": 0.243,
+      "step": 1553
+    },
+    {
+      "epoch": 2.044736842105263,
+      "grad_norm": 0.08012998797966009,
+      "learning_rate": 5.566817737940142e-05,
+      "loss": 0.232,
+      "step": 1554
+    },
+    {
+      "epoch": 2.0460526315789473,
+      "grad_norm": 0.08437656676259887,
+      "learning_rate": 5.553099685984591e-05,
+      "loss": 0.2499,
+      "step": 1555
+    },
+    {
+      "epoch": 2.0473684210526315,
+      "grad_norm": 0.08389268402272403,
+      "learning_rate": 5.5393920572595356e-05,
+      "loss": 0.2318,
+      "step": 1556
+    },
+    {
+      "epoch": 2.0486842105263157,
+      "grad_norm": 0.08099986813273759,
+      "learning_rate": 5.52569488389472e-05,
+      "loss": 0.2408,
+      "step": 1557
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.08017638799598005,
+      "learning_rate": 5.5120081979953785e-05,
+      "loss": 0.2276,
+      "step": 1558
+    },
+    {
+      "epoch": 2.0513157894736844,
+      "grad_norm": 0.08313509302134009,
+      "learning_rate": 5.498332031642177e-05,
+      "loss": 0.2464,
+      "step": 1559
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.08458917823416191,
+      "learning_rate": 5.484666416891109e-05,
+      "loss": 0.2402,
+      "step": 1560
+    },
+    {
+      "epoch": 2.0539473684210527,
+      "grad_norm": 0.08235673729736151,
+      "learning_rate": 5.4710113857734394e-05,
+      "loss": 0.2316,
+      "step": 1561
+    },
+    {
+      "epoch": 2.055263157894737,
+      "grad_norm": 0.08137519978808737,
+      "learning_rate": 5.457366970295634e-05,
+      "loss": 0.2354,
+      "step": 1562
+    },
+    {
+      "epoch": 2.056578947368421,
+      "grad_norm": 0.08503575588624475,
+      "learning_rate": 5.4437332024392694e-05,
+      "loss": 0.2356,
+      "step": 1563
+    },
+    {
+      "epoch": 2.057894736842105,
+      "grad_norm": 0.08267486713807648,
+      "learning_rate": 5.430110114160964e-05,
+      "loss": 0.2492,
+      "step": 1564
+    },
+    {
+      "epoch": 2.0592105263157894,
+      "grad_norm": 0.08202736940009933,
+      "learning_rate": 5.416497737392308e-05,
+      "loss": 0.2418,
+      "step": 1565
+    },
+    {
+      "epoch": 2.0605263157894735,
+      "grad_norm": 0.0848040424518247,
+      "learning_rate": 5.402896104039776e-05,
+      "loss": 0.2385,
+      "step": 1566
+    },
+    {
+      "epoch": 2.0618421052631577,
+      "grad_norm": 0.08202439572712068,
+      "learning_rate": 5.389305245984675e-05,
+      "loss": 0.2411,
+      "step": 1567
+    },
+    {
+      "epoch": 2.0631578947368423,
+      "grad_norm": 0.08444110111591935,
+      "learning_rate": 5.375725195083046e-05,
+      "loss": 0.2365,
+      "step": 1568
+    },
+    {
+      "epoch": 2.0644736842105265,
+      "grad_norm": 0.08419157425499689,
+      "learning_rate": 5.362155983165594e-05,
+      "loss": 0.2472,
+      "step": 1569
+    },
+    {
+      "epoch": 2.0657894736842106,
+      "grad_norm": 0.08359202979261496,
+      "learning_rate": 5.3485976420376336e-05,
+      "loss": 0.2495,
+      "step": 1570
+    },
+    {
+      "epoch": 2.067105263157895,
+      "grad_norm": 0.08127010161214182,
+      "learning_rate": 5.335050203478988e-05,
+      "loss": 0.2319,
+      "step": 1571
+    },
+    {
+      "epoch": 2.068421052631579,
+      "grad_norm": 0.08168717674226678,
+      "learning_rate": 5.321513699243924e-05,
+      "loss": 0.2339,
+      "step": 1572
+    },
+    {
+      "epoch": 2.069736842105263,
+      "grad_norm": 0.08449279952945175,
+      "learning_rate": 5.307988161061085e-05,
+      "loss": 0.238,
+      "step": 1573
+    },
+    {
+      "epoch": 2.0710526315789473,
+      "grad_norm": 0.08225377668981279,
+      "learning_rate": 5.2944736206334034e-05,
+      "loss": 0.2409,
+      "step": 1574
+    },
+    {
+      "epoch": 2.0723684210526314,
+      "grad_norm": 0.0843365596375047,
+      "learning_rate": 5.280970109638047e-05,
+      "loss": 0.246,
+      "step": 1575
+    },
+    {
+      "epoch": 2.0736842105263156,
+      "grad_norm": 0.08534819131960646,
+      "learning_rate": 5.2674776597263186e-05,
+      "loss": 0.2487,
+      "step": 1576
+    },
+    {
+      "epoch": 2.075,
+      "grad_norm": 0.08083479176795237,
+      "learning_rate": 5.253996302523596e-05,
+      "loss": 0.2432,
+      "step": 1577
+    },
+    {
+      "epoch": 2.0763157894736843,
+      "grad_norm": 0.0813017134645149,
+      "learning_rate": 5.240526069629265e-05,
+      "loss": 0.2426,
+      "step": 1578
+    },
+    {
+      "epoch": 2.0776315789473685,
+      "grad_norm": 0.08318443053477804,
+      "learning_rate": 5.227066992616629e-05,
+      "loss": 0.2451,
+      "step": 1579
+    },
+    {
+      "epoch": 2.0789473684210527,
+      "grad_norm": 0.0833772227653302,
+      "learning_rate": 5.2136191030328455e-05,
+      "loss": 0.2351,
+      "step": 1580
+    },
+    {
+      "epoch": 2.080263157894737,
+      "grad_norm": 0.08342495654640115,
+      "learning_rate": 5.2001824323988455e-05,
+      "loss": 0.232,
+      "step": 1581
+    },
+    {
+      "epoch": 2.081578947368421,
+      "grad_norm": 0.08272035294268948,
+      "learning_rate": 5.1867570122092666e-05,
+      "loss": 0.2383,
+      "step": 1582
+    },
+    {
+      "epoch": 2.082894736842105,
+      "grad_norm": 0.08074034588106882,
+      "learning_rate": 5.173342873932383e-05,
+      "loss": 0.232,
+      "step": 1583
+    },
+    {
+      "epoch": 2.0842105263157893,
+      "grad_norm": 0.08255542214386506,
+      "learning_rate": 5.159940049010015e-05,
+      "loss": 0.2346,
+      "step": 1584
+    },
+    {
+      "epoch": 2.085526315789474,
+      "grad_norm": 0.08454238319499686,
+      "learning_rate": 5.146548568857462e-05,
+      "loss": 0.2421,
+      "step": 1585
+    },
+    {
+      "epoch": 2.086842105263158,
+      "grad_norm": 0.084726345772104,
+      "learning_rate": 5.133168464863449e-05,
+      "loss": 0.2387,
+      "step": 1586
+    },
+    {
+      "epoch": 2.088157894736842,
+      "grad_norm": 0.08212224275163124,
+      "learning_rate": 5.1197997683900214e-05,
+      "loss": 0.2352,
+      "step": 1587
+    },
+    {
+      "epoch": 2.0894736842105264,
+      "grad_norm": 0.08285955543488045,
+      "learning_rate": 5.106442510772489e-05,
+      "loss": 0.2246,
+      "step": 1588
+    },
+    {
+      "epoch": 2.0907894736842105,
+      "grad_norm": 0.08221363824415492,
+      "learning_rate": 5.0930967233193504e-05,
+      "loss": 0.2313,
+      "step": 1589
+    },
+    {
+      "epoch": 2.0921052631578947,
+      "grad_norm": 0.08240239248129916,
+      "learning_rate": 5.079762437312219e-05,
+      "loss": 0.2374,
+      "step": 1590
+    },
+    {
+      "epoch": 2.093421052631579,
+      "grad_norm": 0.0833112285945494,
+      "learning_rate": 5.066439684005755e-05,
+      "loss": 0.2354,
+      "step": 1591
+    },
+    {
+      "epoch": 2.094736842105263,
+      "grad_norm": 0.08005461008140058,
+      "learning_rate": 5.0531284946275784e-05,
+      "loss": 0.215,
+      "step": 1592
+    },
+    {
+      "epoch": 2.096052631578947,
+      "grad_norm": 0.08221155288324693,
+      "learning_rate": 5.039828900378204e-05,
+      "loss": 0.2298,
+      "step": 1593
+    },
+    {
+      "epoch": 2.0973684210526318,
+      "grad_norm": 0.08430361712412235,
+      "learning_rate": 5.02654093243098e-05,
+      "loss": 0.2391,
+      "step": 1594
+    },
+    {
+      "epoch": 2.098684210526316,
+      "grad_norm": 0.08646135396687035,
+      "learning_rate": 5.013264621931991e-05,
+      "loss": 0.2382,
+      "step": 1595
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.08349700203559438,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 0.2353,
+      "step": 1596
+    },
+    {
+      "epoch": 2.1013157894736842,
+      "grad_norm": 0.08425984836395248,
+      "learning_rate": 4.986747097726381e-05,
+      "loss": 0.2298,
+      "step": 1597
+    },
+    {
+      "epoch": 2.1026315789473684,
+      "grad_norm": 0.08389516910120101,
+      "learning_rate": 4.97350594617502e-05,
+      "loss": 0.237,
+      "step": 1598
+    },
+    {
+      "epoch": 2.1039473684210526,
+      "grad_norm": 0.08297711753798366,
+      "learning_rate": 4.960276576382283e-05,
+      "loss": 0.2388,
+      "step": 1599
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.08353450960448826,
+      "learning_rate": 4.9470590193569044e-05,
+      "loss": 0.2448,
+      "step": 1600
+    },
+    {
+      "epoch": 2.106578947368421,
+      "grad_norm": 0.08489883369548581,
+      "learning_rate": 4.9338533060799306e-05,
+      "loss": 0.2468,
+      "step": 1601
+    },
+    {
+      "epoch": 2.1078947368421055,
+      "grad_norm": 0.08015122761732316,
+      "learning_rate": 4.920659467504659e-05,
+      "loss": 0.2311,
+      "step": 1602
+    },
+    {
+      "epoch": 2.1092105263157896,
+      "grad_norm": 0.0825490688742884,
+      "learning_rate": 4.907477534556542e-05,
+      "loss": 0.237,
+      "step": 1603
+    },
+    {
+      "epoch": 2.110526315789474,
+      "grad_norm": 0.08391033769643125,
+      "learning_rate": 4.894307538133129e-05,
+      "loss": 0.2384,
+      "step": 1604
+    },
+    {
+      "epoch": 2.111842105263158,
+      "grad_norm": 0.08420377491425757,
+      "learning_rate": 4.8811495091039926e-05,
+      "loss": 0.2365,
+      "step": 1605
+    },
+    {
+      "epoch": 2.113157894736842,
+      "grad_norm": 0.08279227752642704,
+      "learning_rate": 4.868003478310651e-05,
+      "loss": 0.2335,
+      "step": 1606
+    },
+    {
+      "epoch": 2.1144736842105263,
+      "grad_norm": 0.08322872043307299,
+      "learning_rate": 4.854869476566508e-05,
+      "loss": 0.2309,
+      "step": 1607
+    },
+    {
+      "epoch": 2.1157894736842104,
+      "grad_norm": 0.08569452755597119,
+      "learning_rate": 4.841747534656763e-05,
+      "loss": 0.2365,
+      "step": 1608
+    },
+    {
+      "epoch": 2.1171052631578946,
+      "grad_norm": 0.08195650643162714,
+      "learning_rate": 4.828637683338347e-05,
+      "loss": 0.2304,
+      "step": 1609
+    },
+    {
+      "epoch": 2.1184210526315788,
+      "grad_norm": 0.08713567384955483,
+      "learning_rate": 4.815539953339865e-05,
+      "loss": 0.2411,
+      "step": 1610
+    },
+    {
+      "epoch": 2.1197368421052634,
+      "grad_norm": 0.08545867868134706,
+      "learning_rate": 4.802454375361495e-05,
+      "loss": 0.2418,
+      "step": 1611
+    },
+    {
+      "epoch": 2.1210526315789475,
+      "grad_norm": 0.08523598311058483,
+      "learning_rate": 4.7893809800749403e-05,
+      "loss": 0.2328,
+      "step": 1612
+    },
+    {
+      "epoch": 2.1223684210526317,
+      "grad_norm": 0.08672971700799055,
+      "learning_rate": 4.776319798123344e-05,
+      "loss": 0.2357,
+      "step": 1613
+    },
+    {
+      "epoch": 2.123684210526316,
+      "grad_norm": 0.08589391450643363,
+      "learning_rate": 4.763270860121222e-05,
+      "loss": 0.2376,
+      "step": 1614
+    },
+    {
+      "epoch": 2.125,
+      "grad_norm": 0.08657771324163321,
+      "learning_rate": 4.7502341966544e-05,
+      "loss": 0.2409,
+      "step": 1615
+    },
+    {
+      "epoch": 2.126315789473684,
+      "grad_norm": 0.08553999858199614,
+      "learning_rate": 4.737209838279922e-05,
+      "loss": 0.2498,
+      "step": 1616
+    },
+    {
+      "epoch": 2.1276315789473683,
+      "grad_norm": 0.08346189155987148,
+      "learning_rate": 4.7241978155259925e-05,
+      "loss": 0.2382,
+      "step": 1617
+    },
+    {
+      "epoch": 2.1289473684210525,
+      "grad_norm": 0.0832168506244348,
+      "learning_rate": 4.7111981588919084e-05,
+      "loss": 0.2362,
+      "step": 1618
+    },
+    {
+      "epoch": 2.1302631578947366,
+      "grad_norm": 0.08400154849121748,
+      "learning_rate": 4.698210898847976e-05,
+      "loss": 0.2397,
+      "step": 1619
+    },
+    {
+      "epoch": 2.1315789473684212,
+      "grad_norm": 0.08285572712672996,
+      "learning_rate": 4.685236065835443e-05,
+      "loss": 0.2385,
+      "step": 1620
+    },
+    {
+      "epoch": 2.1328947368421054,
+      "grad_norm": 0.08317759957463747,
+      "learning_rate": 4.6722736902664334e-05,
+      "loss": 0.2406,
+      "step": 1621
+    },
+    {
+      "epoch": 2.1342105263157896,
+      "grad_norm": 0.08245858802078364,
+      "learning_rate": 4.659323802523864e-05,
+      "loss": 0.2391,
+      "step": 1622
+    },
+    {
+      "epoch": 2.1355263157894737,
+      "grad_norm": 0.08197930733916409,
+      "learning_rate": 4.646386432961396e-05,
+      "loss": 0.2319,
+      "step": 1623
+    },
+    {
+      "epoch": 2.136842105263158,
+      "grad_norm": 0.08493589285679196,
+      "learning_rate": 4.6334616119033356e-05,
+      "loss": 0.243,
+      "step": 1624
+    },
+    {
+      "epoch": 2.138157894736842,
+      "grad_norm": 0.08490735296967258,
+      "learning_rate": 4.6205493696445754e-05,
+      "loss": 0.236,
+      "step": 1625
+    },
+    {
+      "epoch": 2.139473684210526,
+      "grad_norm": 0.08540155245746917,
+      "learning_rate": 4.6076497364505386e-05,
+      "loss": 0.2517,
+      "step": 1626
+    },
+    {
+      "epoch": 2.1407894736842104,
+      "grad_norm": 0.08620766196769589,
+      "learning_rate": 4.594762742557078e-05,
+      "loss": 0.2317,
+      "step": 1627
+    },
+    {
+      "epoch": 2.1421052631578945,
+      "grad_norm": 0.08354790535329532,
+      "learning_rate": 4.5818884181704294e-05,
+      "loss": 0.2265,
+      "step": 1628
+    },
+    {
+      "epoch": 2.143421052631579,
+      "grad_norm": 0.08468876926659806,
+      "learning_rate": 4.569026793467126e-05,
+      "loss": 0.2323,
+      "step": 1629
+    },
+    {
+      "epoch": 2.1447368421052633,
+      "grad_norm": 0.08422922304528441,
+      "learning_rate": 4.5561778985939366e-05,
+      "loss": 0.231,
+      "step": 1630
+    },
+    {
+      "epoch": 2.1460526315789474,
+      "grad_norm": 0.08597404500241558,
+      "learning_rate": 4.543341763667799e-05,
+      "loss": 0.2455,
+      "step": 1631
+    },
+    {
+      "epoch": 2.1473684210526316,
+      "grad_norm": 0.08569530326388924,
+      "learning_rate": 4.530518418775733e-05,
+      "loss": 0.2374,
+      "step": 1632
+    },
+    {
+      "epoch": 2.1486842105263158,
+      "grad_norm": 0.08651611173999371,
+      "learning_rate": 4.5177078939747796e-05,
+      "loss": 0.2491,
+      "step": 1633
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.08435793058285544,
+      "learning_rate": 4.50491021929194e-05,
+      "loss": 0.2414,
+      "step": 1634
+    },
+    {
+      "epoch": 2.151315789473684,
+      "grad_norm": 0.08452273206770455,
+      "learning_rate": 4.492125424724086e-05,
+      "loss": 0.2389,
+      "step": 1635
+    },
+    {
+      "epoch": 2.1526315789473682,
+      "grad_norm": 0.08400832223995384,
+      "learning_rate": 4.479353540237903e-05,
+      "loss": 0.2509,
+      "step": 1636
+    },
+    {
+      "epoch": 2.153947368421053,
+      "grad_norm": 0.0864191829115749,
+      "learning_rate": 4.466594595769814e-05,
+      "loss": 0.2546,
+      "step": 1637
+    },
+    {
+      "epoch": 2.155263157894737,
+      "grad_norm": 0.08469120634147317,
+      "learning_rate": 4.453848621225912e-05,
+      "loss": 0.2357,
+      "step": 1638
+    },
+    {
+      "epoch": 2.156578947368421,
+      "grad_norm": 0.08564901017872756,
+      "learning_rate": 4.441115646481896e-05,
+      "loss": 0.2367,
+      "step": 1639
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.08581568319788428,
+      "learning_rate": 4.4283957013829846e-05,
+      "loss": 0.238,
+      "step": 1640
+    },
+    {
+      "epoch": 2.1592105263157895,
+      "grad_norm": 0.08280728226775676,
+      "learning_rate": 4.415688815743858e-05,
+      "loss": 0.2296,
+      "step": 1641
+    },
+    {
+      "epoch": 2.1605263157894736,
+      "grad_norm": 0.0853460546951434,
+      "learning_rate": 4.402995019348595e-05,
+      "loss": 0.2387,
+      "step": 1642
+    },
+    {
+      "epoch": 2.161842105263158,
+      "grad_norm": 0.0866721985482825,
+      "learning_rate": 4.390314341950581e-05,
+      "loss": 0.2425,
+      "step": 1643
+    },
+    {
+      "epoch": 2.163157894736842,
+      "grad_norm": 0.08725371319117803,
+      "learning_rate": 4.3776468132724604e-05,
+      "loss": 0.242,
+      "step": 1644
+    },
+    {
+      "epoch": 2.1644736842105265,
+      "grad_norm": 0.08513410768610372,
+      "learning_rate": 4.3649924630060534e-05,
+      "loss": 0.2458,
+      "step": 1645
+    },
+    {
+      "epoch": 2.1657894736842107,
+      "grad_norm": 0.08350902892663929,
+      "learning_rate": 4.35235132081229e-05,
+      "loss": 0.2373,
+      "step": 1646
+    },
+    {
+      "epoch": 2.167105263157895,
+      "grad_norm": 0.08213110851604064,
+      "learning_rate": 4.3397234163211483e-05,
+      "loss": 0.2244,
+      "step": 1647
+    },
+    {
+      "epoch": 2.168421052631579,
+      "grad_norm": 0.08674873213326942,
+      "learning_rate": 4.3271087791315734e-05,
+      "loss": 0.2425,
+      "step": 1648
+    },
+    {
+      "epoch": 2.169736842105263,
+      "grad_norm": 0.08588867781512083,
+      "learning_rate": 4.314507438811407e-05,
+      "loss": 0.2368,
+      "step": 1649
+    },
+    {
+      "epoch": 2.1710526315789473,
+      "grad_norm": 0.08387668032160955,
+      "learning_rate": 4.301919424897338e-05,
+      "loss": 0.2397,
+      "step": 1650
+    },
+    {
+      "epoch": 2.1723684210526315,
+      "grad_norm": 0.08260437607051827,
+      "learning_rate": 4.289344766894807e-05,
+      "loss": 0.2323,
+      "step": 1651
+    },
+    {
+      "epoch": 2.1736842105263157,
+      "grad_norm": 0.0833246040235028,
+      "learning_rate": 4.276783494277954e-05,
+      "loss": 0.2313,
+      "step": 1652
+    },
+    {
+      "epoch": 2.175,
+      "grad_norm": 0.08291199475859631,
+      "learning_rate": 4.264235636489542e-05,
+      "loss": 0.2342,
+      "step": 1653
+    },
+    {
+      "epoch": 2.1763157894736844,
+      "grad_norm": 0.08695364030640872,
+      "learning_rate": 4.2517012229408905e-05,
+      "loss": 0.2534,
+      "step": 1654
+    },
+    {
+      "epoch": 2.1776315789473686,
+      "grad_norm": 0.08428816275057385,
+      "learning_rate": 4.2391802830118135e-05,
+      "loss": 0.2346,
+      "step": 1655
+    },
+    {
+      "epoch": 2.1789473684210527,
+      "grad_norm": 0.084207037568309,
+      "learning_rate": 4.2266728460505375e-05,
+      "loss": 0.2381,
+      "step": 1656
+    },
+    {
+      "epoch": 2.180263157894737,
+      "grad_norm": 0.0886796758669413,
+      "learning_rate": 4.2141789413736354e-05,
+      "loss": 0.2495,
+      "step": 1657
+    },
+    {
+      "epoch": 2.181578947368421,
+      "grad_norm": 0.08307977015154046,
+      "learning_rate": 4.201698598265973e-05,
+      "loss": 0.2327,
+      "step": 1658
+    },
+    {
+      "epoch": 2.182894736842105,
+      "grad_norm": 0.08244618117247395,
+      "learning_rate": 4.189231845980618e-05,
+      "loss": 0.2344,
+      "step": 1659
+    },
+    {
+      "epoch": 2.1842105263157894,
+      "grad_norm": 0.08471751852248358,
+      "learning_rate": 4.176778713738787e-05,
+      "loss": 0.2341,
+      "step": 1660
+    },
+    {
+      "epoch": 2.1855263157894735,
+      "grad_norm": 0.08538574209457075,
+      "learning_rate": 4.164339230729771e-05,
+      "loss": 0.2375,
+      "step": 1661
+    },
+    {
+      "epoch": 2.1868421052631577,
+      "grad_norm": 0.08739644816596194,
+      "learning_rate": 4.151913426110864e-05,
+      "loss": 0.2357,
+      "step": 1662
+    },
+    {
+      "epoch": 2.1881578947368423,
+      "grad_norm": 0.08456784004962821,
+      "learning_rate": 4.13950132900731e-05,
+      "loss": 0.2389,
+      "step": 1663
+    },
+    {
+      "epoch": 2.1894736842105265,
+      "grad_norm": 0.08259644860196363,
+      "learning_rate": 4.127102968512214e-05,
+      "loss": 0.2335,
+      "step": 1664
+    },
+    {
+      "epoch": 2.1907894736842106,
+      "grad_norm": 0.08038218499706021,
+      "learning_rate": 4.114718373686481e-05,
+      "loss": 0.233,
+      "step": 1665
+    },
+    {
+      "epoch": 2.192105263157895,
+      "grad_norm": 0.08297061732865985,
+      "learning_rate": 4.102347573558763e-05,
+      "loss": 0.226,
+      "step": 1666
+    },
+    {
+      "epoch": 2.193421052631579,
+      "grad_norm": 0.08863091885047723,
+      "learning_rate": 4.089990597125368e-05,
+      "loss": 0.2438,
+      "step": 1667
+    },
+    {
+      "epoch": 2.194736842105263,
+      "grad_norm": 0.08628873633777266,
+      "learning_rate": 4.077647473350201e-05,
+      "loss": 0.2361,
+      "step": 1668
+    },
+    {
+      "epoch": 2.1960526315789473,
+      "grad_norm": 0.08360640948058073,
+      "learning_rate": 4.065318231164704e-05,
+      "loss": 0.2417,
+      "step": 1669
+    },
+    {
+      "epoch": 2.1973684210526314,
+      "grad_norm": 0.08524943894134786,
+      "learning_rate": 4.053002899467774e-05,
+      "loss": 0.2387,
+      "step": 1670
+    },
+    {
+      "epoch": 2.1986842105263156,
+      "grad_norm": 0.08537007936359343,
+      "learning_rate": 4.040701507125712e-05,
+      "loss": 0.2408,
+      "step": 1671
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.08300365174925911,
+      "learning_rate": 4.028414082972141e-05,
+      "loss": 0.226,
+      "step": 1672
+    },
+    {
+      "epoch": 2.2013157894736843,
+      "grad_norm": 0.08790060045013368,
+      "learning_rate": 4.016140655807936e-05,
+      "loss": 0.2507,
+      "step": 1673
+    },
+    {
+      "epoch": 2.2026315789473685,
+      "grad_norm": 0.0831197113798778,
+      "learning_rate": 4.003881254401183e-05,
+      "loss": 0.2343,
+      "step": 1674
+    },
+    {
+      "epoch": 2.2039473684210527,
+      "grad_norm": 0.08118714142897289,
+      "learning_rate": 3.991635907487076e-05,
+      "loss": 0.2265,
+      "step": 1675
+    },
+    {
+      "epoch": 2.205263157894737,
+      "grad_norm": 0.08836758200078478,
+      "learning_rate": 3.97940464376787e-05,
+      "loss": 0.2539,
+      "step": 1676
+    },
+    {
+      "epoch": 2.206578947368421,
+      "grad_norm": 0.0829126388696676,
+      "learning_rate": 3.967187491912813e-05,
+      "loss": 0.2343,
+      "step": 1677
+    },
+    {
+      "epoch": 2.207894736842105,
+      "grad_norm": 0.08050172289656292,
+      "learning_rate": 3.9549844805580706e-05,
+      "loss": 0.2339,
+      "step": 1678
+    },
+    {
+      "epoch": 2.2092105263157893,
+      "grad_norm": 0.08394767599321681,
+      "learning_rate": 3.942795638306674e-05,
+      "loss": 0.2369,
+      "step": 1679
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.08711738632474018,
+      "learning_rate": 3.9306209937284346e-05,
+      "loss": 0.2419,
+      "step": 1680
+    },
+    {
+      "epoch": 2.211842105263158,
+      "grad_norm": 0.0814409575586451,
+      "learning_rate": 3.918460575359882e-05,
+      "loss": 0.2306,
+      "step": 1681
+    },
+    {
+      "epoch": 2.213157894736842,
+      "grad_norm": 0.08526205771283932,
+      "learning_rate": 3.906314411704215e-05,
+      "loss": 0.2374,
+      "step": 1682
+    },
+    {
+      "epoch": 2.2144736842105264,
+      "grad_norm": 0.08590081960012944,
+      "learning_rate": 3.8941825312312054e-05,
+      "loss": 0.2324,
+      "step": 1683
+    },
+    {
+      "epoch": 2.2157894736842105,
+      "grad_norm": 0.08577840811573154,
+      "learning_rate": 3.882064962377154e-05,
+      "loss": 0.2402,
+      "step": 1684
+    },
+    {
+      "epoch": 2.2171052631578947,
+      "grad_norm": 0.08559702425445263,
+      "learning_rate": 3.869961733544814e-05,
+      "loss": 0.2402,
+      "step": 1685
+    },
+    {
+      "epoch": 2.218421052631579,
+      "grad_norm": 0.08436964025823426,
+      "learning_rate": 3.857872873103322e-05,
+      "loss": 0.2415,
+      "step": 1686
+    },
+    {
+      "epoch": 2.219736842105263,
+      "grad_norm": 0.08396240106431019,
+      "learning_rate": 3.845798409388149e-05,
+      "loss": 0.2302,
+      "step": 1687
+    },
+    {
+      "epoch": 2.221052631578947,
+      "grad_norm": 0.08629968011623863,
+      "learning_rate": 3.83373837070101e-05,
+      "loss": 0.2358,
+      "step": 1688
+    },
+    {
+      "epoch": 2.2223684210526318,
+      "grad_norm": 0.0838099558892786,
+      "learning_rate": 3.821692785309807e-05,
+      "loss": 0.2334,
+      "step": 1689
+    },
+    {
+      "epoch": 2.223684210526316,
+      "grad_norm": 0.08557035987280985,
+      "learning_rate": 3.809661681448576e-05,
+      "loss": 0.2311,
+      "step": 1690
+    },
+    {
+      "epoch": 2.225,
+      "grad_norm": 0.08591483983457045,
+      "learning_rate": 3.7976450873174005e-05,
+      "loss": 0.2421,
+      "step": 1691
+    },
+    {
+      "epoch": 2.2263157894736842,
+      "grad_norm": 0.08701077906025538,
+      "learning_rate": 3.7856430310823545e-05,
+      "loss": 0.2451,
+      "step": 1692
+    },
+    {
+      "epoch": 2.2276315789473684,
+      "grad_norm": 0.08152777587376067,
+      "learning_rate": 3.773655540875438e-05,
+      "loss": 0.2335,
+      "step": 1693
+    },
+    {
+      "epoch": 2.2289473684210526,
+      "grad_norm": 0.0879771188843296,
+      "learning_rate": 3.7616826447945066e-05,
+      "loss": 0.2404,
+      "step": 1694
+    },
+    {
+      "epoch": 2.2302631578947367,
+      "grad_norm": 0.08415415636312946,
+      "learning_rate": 3.749724370903216e-05,
+      "loss": 0.2355,
+      "step": 1695
+    },
+    {
+      "epoch": 2.231578947368421,
+      "grad_norm": 0.08756860364801405,
+      "learning_rate": 3.737780747230941e-05,
+      "loss": 0.2378,
+      "step": 1696
+    },
+    {
+      "epoch": 2.2328947368421055,
+      "grad_norm": 0.08454219628982126,
+      "learning_rate": 3.725851801772715e-05,
+      "loss": 0.2445,
+      "step": 1697
+    },
+    {
+      "epoch": 2.2342105263157896,
+      "grad_norm": 0.0869901636562866,
+      "learning_rate": 3.713937562489179e-05,
+      "loss": 0.2296,
+      "step": 1698
+    },
+    {
+      "epoch": 2.235526315789474,
+      "grad_norm": 0.08630652964552962,
+      "learning_rate": 3.702038057306492e-05,
+      "loss": 0.2302,
+      "step": 1699
+    },
+    {
+      "epoch": 2.236842105263158,
+      "grad_norm": 0.08659038929154979,
+      "learning_rate": 3.69015331411628e-05,
+      "loss": 0.2351,
+      "step": 1700
+    },
+    {
+      "epoch": 2.238157894736842,
+      "grad_norm": 0.08649330336962248,
+      "learning_rate": 3.678283360775571e-05,
+      "loss": 0.2362,
+      "step": 1701
+    },
+    {
+      "epoch": 2.2394736842105263,
+      "grad_norm": 0.08424707415501553,
+      "learning_rate": 3.6664282251067184e-05,
+      "loss": 0.2378,
+      "step": 1702
+    },
+    {
+      "epoch": 2.2407894736842104,
+      "grad_norm": 0.08338435013878512,
+      "learning_rate": 3.65458793489736e-05,
+      "loss": 0.2336,
+      "step": 1703
+    },
+    {
+      "epoch": 2.2421052631578946,
+      "grad_norm": 0.08719248023927356,
+      "learning_rate": 3.642762517900322e-05,
+      "loss": 0.2426,
+      "step": 1704
+    },
+    {
+      "epoch": 2.2434210526315788,
+      "grad_norm": 0.08436003020889046,
+      "learning_rate": 3.6309520018335705e-05,
+      "loss": 0.2301,
+      "step": 1705
+    },
+    {
+      "epoch": 2.2447368421052634,
+      "grad_norm": 0.08728623567451708,
+      "learning_rate": 3.619156414380156e-05,
+      "loss": 0.2462,
+      "step": 1706
+    },
+    {
+      "epoch": 2.2460526315789475,
+      "grad_norm": 0.08273816077006978,
+      "learning_rate": 3.607375783188125e-05,
+      "loss": 0.2433,
+      "step": 1707
+    },
+    {
+      "epoch": 2.2473684210526317,
+      "grad_norm": 0.08380037356624727,
+      "learning_rate": 3.595610135870472e-05,
+      "loss": 0.2419,
+      "step": 1708
+    },
+    {
+      "epoch": 2.248684210526316,
+      "grad_norm": 0.08599999255981298,
+      "learning_rate": 3.583859500005071e-05,
+      "loss": 0.2459,
+      "step": 1709
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.0857928049153299,
+      "learning_rate": 3.5721239031346066e-05,
+      "loss": 0.2332,
+      "step": 1710
+    },
+    {
+      "epoch": 2.251315789473684,
+      "grad_norm": 0.08563816332335225,
+      "learning_rate": 3.560403372766522e-05,
+      "loss": 0.239,
+      "step": 1711
+    },
+    {
+      "epoch": 2.2526315789473683,
+      "grad_norm": 0.08648184240333989,
+      "learning_rate": 3.548697936372937e-05,
+      "loss": 0.2369,
+      "step": 1712
+    },
+    {
+      "epoch": 2.2539473684210525,
+      "grad_norm": 0.08405093979550926,
+      "learning_rate": 3.53700762139059e-05,
+      "loss": 0.2484,
+      "step": 1713
+    },
+    {
+      "epoch": 2.2552631578947366,
+      "grad_norm": 0.08629123371639234,
+      "learning_rate": 3.525332455220789e-05,
+      "loss": 0.2526,
+      "step": 1714
+    },
+    {
+      "epoch": 2.2565789473684212,
+      "grad_norm": 0.08603955512267374,
+      "learning_rate": 3.5136724652293206e-05,
+      "loss": 0.2363,
+      "step": 1715
+    },
+    {
+      "epoch": 2.2578947368421054,
+      "grad_norm": 0.08357218310355902,
+      "learning_rate": 3.5020276787464056e-05,
+      "loss": 0.2379,
+      "step": 1716
+    },
+    {
+      "epoch": 2.2592105263157896,
+      "grad_norm": 0.08178687870308674,
+      "learning_rate": 3.490398123066628e-05,
+      "loss": 0.2342,
+      "step": 1717
+    },
+    {
+      "epoch": 2.2605263157894737,
+      "grad_norm": 0.08618629426674392,
+      "learning_rate": 3.4787838254488694e-05,
+      "loss": 0.2436,
+      "step": 1718
+    },
+    {
+      "epoch": 2.261842105263158,
+      "grad_norm": 0.08654678035853532,
+      "learning_rate": 3.4671848131162544e-05,
+      "loss": 0.2413,
+      "step": 1719
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.08154103451407675,
+      "learning_rate": 3.455601113256073e-05,
+      "loss": 0.2445,
+      "step": 1720
+    },
+    {
+      "epoch": 2.264473684210526,
+      "grad_norm": 0.08283767549879928,
+      "learning_rate": 3.444032753019723e-05,
+      "loss": 0.235,
+      "step": 1721
+    },
+    {
+      "epoch": 2.2657894736842104,
+      "grad_norm": 0.08402970526889068,
+      "learning_rate": 3.4324797595226565e-05,
+      "loss": 0.2337,
+      "step": 1722
+    },
+    {
+      "epoch": 2.2671052631578945,
+      "grad_norm": 0.08237791548613392,
+      "learning_rate": 3.420942159844298e-05,
+      "loss": 0.2352,
+      "step": 1723
+    },
+    {
+      "epoch": 2.268421052631579,
+      "grad_norm": 0.08344712694132657,
+      "learning_rate": 3.4094199810279924e-05,
+      "loss": 0.236,
+      "step": 1724
+    },
+    {
+      "epoch": 2.2697368421052633,
+      "grad_norm": 0.08799849742749684,
+      "learning_rate": 3.3979132500809405e-05,
+      "loss": 0.2481,
+      "step": 1725
+    },
+    {
+      "epoch": 2.2710526315789474,
+      "grad_norm": 0.08588090804811839,
+      "learning_rate": 3.386421993974129e-05,
+      "loss": 0.2364,
+      "step": 1726
+    },
+    {
+      "epoch": 2.2723684210526316,
+      "grad_norm": 0.0838270871173491,
+      "learning_rate": 3.3749462396422846e-05,
+      "loss": 0.228,
+      "step": 1727
+    },
+    {
+      "epoch": 2.2736842105263158,
+      "grad_norm": 0.08845792107580112,
+      "learning_rate": 3.363486013983788e-05,
+      "loss": 0.2437,
+      "step": 1728
+    },
+    {
+      "epoch": 2.275,
+      "grad_norm": 0.08775832799442573,
+      "learning_rate": 3.352041343860621e-05,
+      "loss": 0.2433,
+      "step": 1729
+    },
+    {
+      "epoch": 2.276315789473684,
+      "grad_norm": 0.08852510695897718,
+      "learning_rate": 3.340612256098316e-05,
+      "loss": 0.2528,
+      "step": 1730
+    },
+    {
+      "epoch": 2.2776315789473682,
+      "grad_norm": 0.09072365256400772,
+      "learning_rate": 3.329198777485869e-05,
+      "loss": 0.235,
+      "step": 1731
+    },
+    {
+      "epoch": 2.2789473684210524,
+      "grad_norm": 0.08478552332865018,
+      "learning_rate": 3.317800934775696e-05,
+      "loss": 0.2419,
+      "step": 1732
+    },
+    {
+      "epoch": 2.280263157894737,
+      "grad_norm": 0.08510050487380844,
+      "learning_rate": 3.30641875468356e-05,
+      "loss": 0.2382,
+      "step": 1733
+    },
+    {
+      "epoch": 2.281578947368421,
+      "grad_norm": 0.08450032284933855,
+      "learning_rate": 3.2950522638885106e-05,
+      "loss": 0.2411,
+      "step": 1734
+    },
+    {
+      "epoch": 2.2828947368421053,
+      "grad_norm": 0.0847645250276902,
+      "learning_rate": 3.283701489032832e-05,
+      "loss": 0.2289,
+      "step": 1735
+    },
+    {
+      "epoch": 2.2842105263157895,
+      "grad_norm": 0.08500704575747163,
+      "learning_rate": 3.2723664567219626e-05,
+      "loss": 0.2362,
+      "step": 1736
+    },
+    {
+      "epoch": 2.2855263157894736,
+      "grad_norm": 0.08540028460798461,
+      "learning_rate": 3.261047193524439e-05,
+      "loss": 0.2441,
+      "step": 1737
+    },
+    {
+      "epoch": 2.286842105263158,
+      "grad_norm": 0.08482142626182317,
+      "learning_rate": 3.249743725971849e-05,
+      "loss": 0.2339,
+      "step": 1738
+    },
+    {
+      "epoch": 2.288157894736842,
+      "grad_norm": 0.08455145109428938,
+      "learning_rate": 3.238456080558743e-05,
+      "loss": 0.2364,
+      "step": 1739
+    },
+    {
+      "epoch": 2.2894736842105265,
+      "grad_norm": 0.08608434293062465,
+      "learning_rate": 3.227184283742591e-05,
+      "loss": 0.2427,
+      "step": 1740
+    },
+    {
+      "epoch": 2.2907894736842107,
+      "grad_norm": 0.08525626191346178,
+      "learning_rate": 3.2159283619437155e-05,
+      "loss": 0.2447,
+      "step": 1741
+    },
+    {
+      "epoch": 2.292105263157895,
+      "grad_norm": 0.0867443524413285,
+      "learning_rate": 3.2046883415452246e-05,
+      "loss": 0.2297,
+      "step": 1742
+    },
+    {
+      "epoch": 2.293421052631579,
+      "grad_norm": 0.08427294151144223,
+      "learning_rate": 3.193464248892964e-05,
+      "loss": 0.2466,
+      "step": 1743
+    },
+    {
+      "epoch": 2.294736842105263,
+      "grad_norm": 0.0849042097519616,
+      "learning_rate": 3.182256110295437e-05,
+      "loss": 0.2402,
+      "step": 1744
+    },
+    {
+      "epoch": 2.2960526315789473,
+      "grad_norm": 0.08773216559618434,
+      "learning_rate": 3.171063952023753e-05,
+      "loss": 0.238,
+      "step": 1745
+    },
+    {
+      "epoch": 2.2973684210526315,
+      "grad_norm": 0.08514375299514274,
+      "learning_rate": 3.159887800311569e-05,
+      "loss": 0.2495,
+      "step": 1746
+    },
+    {
+      "epoch": 2.2986842105263157,
+      "grad_norm": 0.08221174413038916,
+      "learning_rate": 3.148727681355022e-05,
+      "loss": 0.2327,
+      "step": 1747
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.08292212985659095,
+      "learning_rate": 3.137583621312665e-05,
+      "loss": 0.2265,
+      "step": 1748
+    },
+    {
+      "epoch": 2.3013157894736844,
+      "grad_norm": 0.08536339352006025,
+      "learning_rate": 3.126455646305416e-05,
+      "loss": 0.2316,
+      "step": 1749
+    },
+    {
+      "epoch": 2.3026315789473686,
+      "grad_norm": 0.083425418514592,
+      "learning_rate": 3.115343782416483e-05,
+      "loss": 0.2405,
+      "step": 1750
+    },
+    {
+      "epoch": 2.3039473684210527,
+      "grad_norm": 0.08635976351926793,
+      "learning_rate": 3.1042480556913224e-05,
+      "loss": 0.234,
+      "step": 1751
+    },
+    {
+      "epoch": 2.305263157894737,
+      "grad_norm": 0.08648353037146853,
+      "learning_rate": 3.093168492137557e-05,
+      "loss": 0.222,
+      "step": 1752
+    },
+    {
+      "epoch": 2.306578947368421,
+      "grad_norm": 0.08421631985285294,
+      "learning_rate": 3.082105117724923e-05,
+      "loss": 0.2301,
+      "step": 1753
+    },
+    {
+      "epoch": 2.307894736842105,
+      "grad_norm": 0.08776858485770223,
+      "learning_rate": 3.071057958385221e-05,
+      "loss": 0.236,
+      "step": 1754
+    },
+    {
+      "epoch": 2.3092105263157894,
+      "grad_norm": 0.08692453013595573,
+      "learning_rate": 3.0600270400122335e-05,
+      "loss": 0.2379,
+      "step": 1755
+    },
+    {
+      "epoch": 2.3105263157894735,
+      "grad_norm": 0.08408758438701626,
+      "learning_rate": 3.0490123884616796e-05,
+      "loss": 0.2397,
+      "step": 1756
+    },
+    {
+      "epoch": 2.3118421052631577,
+      "grad_norm": 0.08561383977421695,
+      "learning_rate": 3.0380140295511516e-05,
+      "loss": 0.2362,
+      "step": 1757
+    },
+    {
+      "epoch": 2.3131578947368423,
+      "grad_norm": 0.08523323066228368,
+      "learning_rate": 3.0270319890600462e-05,
+      "loss": 0.2373,
+      "step": 1758
+    },
+    {
+      "epoch": 2.3144736842105265,
+      "grad_norm": 0.08384268173924093,
+      "learning_rate": 3.0160662927295225e-05,
+      "loss": 0.2221,
+      "step": 1759
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.08377079729896948,
+      "learning_rate": 3.0051169662624225e-05,
+      "loss": 0.2305,
+      "step": 1760
+    },
+    {
+      "epoch": 2.317105263157895,
+      "grad_norm": 0.08434931689091195,
+      "learning_rate": 2.994184035323213e-05,
+      "loss": 0.2264,
+      "step": 1761
+    },
+    {
+      "epoch": 2.318421052631579,
+      "grad_norm": 0.08485228565126582,
+      "learning_rate": 2.983267525537945e-05,
+      "loss": 0.2407,
+      "step": 1762
+    },
+    {
+      "epoch": 2.319736842105263,
+      "grad_norm": 0.08458549101284824,
+      "learning_rate": 2.9723674624941688e-05,
+      "loss": 0.2397,
+      "step": 1763
+    },
+    {
+      "epoch": 2.3210526315789473,
+      "grad_norm": 0.08376501835808486,
+      "learning_rate": 2.9614838717408867e-05,
+      "loss": 0.2353,
+      "step": 1764
+    },
+    {
+      "epoch": 2.3223684210526314,
+      "grad_norm": 0.08574531246445859,
+      "learning_rate": 2.950616778788492e-05,
+      "loss": 0.2465,
+      "step": 1765
+    },
+    {
+      "epoch": 2.3236842105263156,
+      "grad_norm": 0.08775545996621473,
+      "learning_rate": 2.9397662091087054e-05,
+      "loss": 0.2464,
+      "step": 1766
+    },
+    {
+      "epoch": 2.325,
+      "grad_norm": 0.0818496814028781,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 0.2204,
+      "step": 1767
+    },
+    {
+      "epoch": 2.3263157894736843,
+      "grad_norm": 0.08595948586857366,
+      "learning_rate": 2.9181147412601562e-05,
+      "loss": 0.2329,
+      "step": 1768
+    },
+    {
+      "epoch": 2.3276315789473685,
+      "grad_norm": 0.08376848170186316,
+      "learning_rate": 2.9073138938409495e-05,
+      "loss": 0.2416,
+      "step": 1769
+    },
+    {
+      "epoch": 2.3289473684210527,
+      "grad_norm": 0.08594662597858775,
+      "learning_rate": 2.89652967119336e-05,
+      "loss": 0.2439,
+      "step": 1770
+    },
+    {
+      "epoch": 2.330263157894737,
+      "grad_norm": 0.08548685030438735,
+      "learning_rate": 2.8857620985948652e-05,
+      "loss": 0.2466,
+      "step": 1771
+    },
+    {
+      "epoch": 2.331578947368421,
+      "grad_norm": 0.08440124314126071,
+      "learning_rate": 2.8750112012839214e-05,
+      "loss": 0.2318,
+      "step": 1772
+    },
+    {
+      "epoch": 2.332894736842105,
+      "grad_norm": 0.08563767620958296,
+      "learning_rate": 2.8642770044598966e-05,
+      "loss": 0.2386,
+      "step": 1773
+    },
+    {
+      "epoch": 2.3342105263157893,
+      "grad_norm": 0.08413261927528606,
+      "learning_rate": 2.8535595332830102e-05,
+      "loss": 0.2412,
+      "step": 1774
+    },
+    {
+      "epoch": 2.3355263157894735,
+      "grad_norm": 0.08430887282829173,
+      "learning_rate": 2.8428588128742894e-05,
+      "loss": 0.2433,
+      "step": 1775
+    },
+    {
+      "epoch": 2.336842105263158,
+      "grad_norm": 0.08421978093619426,
+      "learning_rate": 2.8321748683154893e-05,
+      "loss": 0.2266,
+      "step": 1776
+    },
+    {
+      "epoch": 2.338157894736842,
+      "grad_norm": 0.08367626319071061,
+      "learning_rate": 2.8215077246490417e-05,
+      "loss": 0.232,
+      "step": 1777
+    },
+    {
+      "epoch": 2.3394736842105264,
+      "grad_norm": 0.0852480773032482,
+      "learning_rate": 2.810857406878009e-05,
+      "loss": 0.2431,
+      "step": 1778
+    },
+    {
+      "epoch": 2.3407894736842105,
+      "grad_norm": 0.08752279412974256,
+      "learning_rate": 2.800223939966007e-05,
+      "loss": 0.2392,
+      "step": 1779
+    },
+    {
+      "epoch": 2.3421052631578947,
+      "grad_norm": 0.08525332318728705,
+      "learning_rate": 2.789607348837153e-05,
+      "loss": 0.2261,
+      "step": 1780
+    },
+    {
+      "epoch": 2.343421052631579,
+      "grad_norm": 0.08501669465697845,
+      "learning_rate": 2.7790076583760126e-05,
+      "loss": 0.236,
+      "step": 1781
+    },
+    {
+      "epoch": 2.344736842105263,
+      "grad_norm": 0.08586501169000936,
+      "learning_rate": 2.7684248934275325e-05,
+      "loss": 0.2369,
+      "step": 1782
+    },
+    {
+      "epoch": 2.3460526315789476,
+      "grad_norm": 0.08330526797227158,
+      "learning_rate": 2.757859078796997e-05,
+      "loss": 0.2426,
+      "step": 1783
+    },
+    {
+      "epoch": 2.3473684210526318,
+      "grad_norm": 0.08682682618319912,
+      "learning_rate": 2.7473102392499518e-05,
+      "loss": 0.2405,
+      "step": 1784
+    },
+    {
+      "epoch": 2.348684210526316,
+      "grad_norm": 0.08681140965019736,
+      "learning_rate": 2.73677839951215e-05,
+      "loss": 0.2438,
+      "step": 1785
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.08557189396659094,
+      "learning_rate": 2.7262635842695127e-05,
+      "loss": 0.232,
+      "step": 1786
+    },
+    {
+      "epoch": 2.3513157894736842,
+      "grad_norm": 0.08357509250185297,
+      "learning_rate": 2.7157658181680457e-05,
+      "loss": 0.2391,
+      "step": 1787
+    },
+    {
+      "epoch": 2.3526315789473684,
+      "grad_norm": 0.0863037240334069,
+      "learning_rate": 2.7052851258137935e-05,
+      "loss": 0.2385,
+      "step": 1788
+    },
+    {
+      "epoch": 2.3539473684210526,
+      "grad_norm": 0.08781831219541067,
+      "learning_rate": 2.6948215317727844e-05,
+      "loss": 0.2332,
+      "step": 1789
+    },
+    {
+      "epoch": 2.3552631578947367,
+      "grad_norm": 0.08657991059589279,
+      "learning_rate": 2.684375060570965e-05,
+      "loss": 0.2386,
+      "step": 1790
+    },
+    {
+      "epoch": 2.356578947368421,
+      "grad_norm": 0.08748510488863394,
+      "learning_rate": 2.6739457366941543e-05,
+      "loss": 0.2449,
+      "step": 1791
+    },
+    {
+      "epoch": 2.3578947368421055,
+      "grad_norm": 0.08693207168372159,
+      "learning_rate": 2.6635335845879737e-05,
+      "loss": 0.2294,
+      "step": 1792
+    },
+    {
+      "epoch": 2.3592105263157896,
+      "grad_norm": 0.0867096331580623,
+      "learning_rate": 2.653138628657793e-05,
+      "loss": 0.2375,
+      "step": 1793
+    },
+    {
+      "epoch": 2.360526315789474,
+      "grad_norm": 0.09001793260204524,
+      "learning_rate": 2.6427608932686843e-05,
+      "loss": 0.2479,
+      "step": 1794
+    },
+    {
+      "epoch": 2.361842105263158,
+      "grad_norm": 0.08716676015459905,
+      "learning_rate": 2.6324004027453464e-05,
+      "loss": 0.2371,
+      "step": 1795
+    },
+    {
+      "epoch": 2.363157894736842,
+      "grad_norm": 0.08978495624040321,
+      "learning_rate": 2.622057181372063e-05,
+      "loss": 0.2468,
+      "step": 1796
+    },
+    {
+      "epoch": 2.3644736842105263,
+      "grad_norm": 0.08299772256887776,
+      "learning_rate": 2.6117312533926362e-05,
+      "loss": 0.2336,
+      "step": 1797
+    },
+    {
+      "epoch": 2.3657894736842104,
+      "grad_norm": 0.08607109549303381,
+      "learning_rate": 2.601422643010335e-05,
+      "loss": 0.2424,
+      "step": 1798
+    },
+    {
+      "epoch": 2.3671052631578946,
+      "grad_norm": 0.08507498429800535,
+      "learning_rate": 2.5911313743878418e-05,
+      "loss": 0.2385,
+      "step": 1799
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.08326170794330705,
+      "learning_rate": 2.5808574716471856e-05,
+      "loss": 0.2419,
+      "step": 1800
+    },
+    {
+      "epoch": 2.3697368421052634,
+      "grad_norm": 0.08536319126880634,
+      "learning_rate": 2.570600958869689e-05,
+      "loss": 0.2337,
+      "step": 1801
+    },
+    {
+      "epoch": 2.3710526315789475,
+      "grad_norm": 0.0832285855296963,
+      "learning_rate": 2.5603618600959223e-05,
+      "loss": 0.2262,
+      "step": 1802
+    },
+    {
+      "epoch": 2.3723684210526317,
+      "grad_norm": 0.08748256088729627,
+      "learning_rate": 2.55014019932563e-05,
+      "loss": 0.238,
+      "step": 1803
+    },
+    {
+      "epoch": 2.373684210526316,
+      "grad_norm": 0.0845403743408863,
+      "learning_rate": 2.5399360005176886e-05,
+      "loss": 0.2342,
+      "step": 1804
+    },
+    {
+      "epoch": 2.375,
+      "grad_norm": 0.08306478491889778,
+      "learning_rate": 2.529749287590042e-05,
+      "loss": 0.236,
+      "step": 1805
+    },
+    {
+      "epoch": 2.376315789473684,
+      "grad_norm": 0.08506256922386096,
+      "learning_rate": 2.519580084419646e-05,
+      "loss": 0.2364,
+      "step": 1806
+    },
+    {
+      "epoch": 2.3776315789473683,
+      "grad_norm": 0.0821793026467122,
+      "learning_rate": 2.509428414842424e-05,
+      "loss": 0.2241,
+      "step": 1807
+    },
+    {
+      "epoch": 2.3789473684210525,
+      "grad_norm": 0.08556195534896847,
+      "learning_rate": 2.4992943026531935e-05,
+      "loss": 0.2365,
+      "step": 1808
+    },
+    {
+      "epoch": 2.3802631578947366,
+      "grad_norm": 0.08705793209497273,
+      "learning_rate": 2.4891777716056176e-05,
+      "loss": 0.2526,
+      "step": 1809
+    },
+    {
+      "epoch": 2.3815789473684212,
+      "grad_norm": 0.08705293536857067,
+      "learning_rate": 2.4790788454121584e-05,
+      "loss": 0.2361,
+      "step": 1810
+    },
+    {
+      "epoch": 2.3828947368421054,
+      "grad_norm": 0.08488258217639373,
+      "learning_rate": 2.4689975477440086e-05,
+      "loss": 0.2273,
+      "step": 1811
+    },
+    {
+      "epoch": 2.3842105263157896,
+      "grad_norm": 0.09044220228981603,
+      "learning_rate": 2.4589339022310386e-05,
+      "loss": 0.2532,
+      "step": 1812
+    },
+    {
+      "epoch": 2.3855263157894737,
+      "grad_norm": 0.0858040156345174,
+      "learning_rate": 2.4488879324617474e-05,
+      "loss": 0.2395,
+      "step": 1813
+    },
+    {
+      "epoch": 2.386842105263158,
+      "grad_norm": 0.08540339576774533,
+      "learning_rate": 2.4388596619831993e-05,
+      "loss": 0.2393,
+      "step": 1814
+    },
+    {
+      "epoch": 2.388157894736842,
+      "grad_norm": 0.08684868336556462,
+      "learning_rate": 2.4288491143009795e-05,
+      "loss": 0.2362,
+      "step": 1815
+    },
+    {
+      "epoch": 2.389473684210526,
+      "grad_norm": 0.086881482942677,
+      "learning_rate": 2.4188563128791254e-05,
+      "loss": 0.2352,
+      "step": 1816
+    },
+    {
+      "epoch": 2.3907894736842104,
+      "grad_norm": 0.08698301322688715,
+      "learning_rate": 2.4088812811400773e-05,
+      "loss": 0.2366,
+      "step": 1817
+    },
+    {
+      "epoch": 2.3921052631578945,
+      "grad_norm": 0.08860609435080301,
+      "learning_rate": 2.3989240424646355e-05,
+      "loss": 0.2579,
+      "step": 1818
+    },
+    {
+      "epoch": 2.393421052631579,
+      "grad_norm": 0.08501020836673648,
+      "learning_rate": 2.388984620191883e-05,
+      "loss": 0.2305,
+      "step": 1819
+    },
+    {
+      "epoch": 2.3947368421052633,
+      "grad_norm": 0.08510368650070274,
+      "learning_rate": 2.379063037619146e-05,
+      "loss": 0.247,
+      "step": 1820
+    },
+    {
+      "epoch": 2.3960526315789474,
+      "grad_norm": 0.0829406506692919,
+      "learning_rate": 2.3691593180019366e-05,
+      "loss": 0.2278,
+      "step": 1821
+    },
+    {
+      "epoch": 2.3973684210526316,
+      "grad_norm": 0.0867391319954807,
+      "learning_rate": 2.3592734845538956e-05,
+      "loss": 0.2336,
+      "step": 1822
+    },
+    {
+      "epoch": 2.3986842105263158,
+      "grad_norm": 0.08541140823531897,
+      "learning_rate": 2.3494055604467447e-05,
+      "loss": 0.2492,
+      "step": 1823
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.08302602754112574,
+      "learning_rate": 2.339555568810221e-05,
+      "loss": 0.2373,
+      "step": 1824
+    },
+    {
+      "epoch": 2.401315789473684,
+      "grad_norm": 0.08577260892930051,
+      "learning_rate": 2.32972353273203e-05,
+      "loss": 0.2419,
+      "step": 1825
+    },
+    {
+      "epoch": 2.4026315789473682,
+      "grad_norm": 0.08774292891845863,
+      "learning_rate": 2.319909475257799e-05,
+      "loss": 0.2405,
+      "step": 1826
+    },
+    {
+      "epoch": 2.4039473684210524,
+      "grad_norm": 0.08550920787272853,
+      "learning_rate": 2.3101134193910024e-05,
+      "loss": 0.2434,
+      "step": 1827
+    },
+    {
+      "epoch": 2.405263157894737,
+      "grad_norm": 0.08426612023984284,
+      "learning_rate": 2.300335388092929e-05,
+      "loss": 0.2234,
+      "step": 1828
+    },
+    {
+      "epoch": 2.406578947368421,
+      "grad_norm": 0.0848436411883103,
+      "learning_rate": 2.2905754042826143e-05,
+      "loss": 0.2377,
+      "step": 1829
+    },
+    {
+      "epoch": 2.4078947368421053,
+      "grad_norm": 0.08552189669750118,
+      "learning_rate": 2.2808334908367914e-05,
+      "loss": 0.243,
+      "step": 1830
+    },
+    {
+      "epoch": 2.4092105263157895,
+      "grad_norm": 0.08572236225629354,
+      "learning_rate": 2.271109670589844e-05,
+      "loss": 0.2398,
+      "step": 1831
+    },
+    {
+      "epoch": 2.4105263157894736,
+      "grad_norm": 0.08511556223223665,
+      "learning_rate": 2.2614039663337417e-05,
+      "loss": 0.225,
+      "step": 1832
+    },
+    {
+      "epoch": 2.411842105263158,
+      "grad_norm": 0.08510989348259673,
+      "learning_rate": 2.2517164008179882e-05,
+      "loss": 0.2393,
+      "step": 1833
+    },
+    {
+      "epoch": 2.413157894736842,
+      "grad_norm": 0.08621647138074637,
+      "learning_rate": 2.2420469967495793e-05,
+      "loss": 0.2492,
+      "step": 1834
+    },
+    {
+      "epoch": 2.4144736842105265,
+      "grad_norm": 0.08627144976057537,
+      "learning_rate": 2.232395776792938e-05,
+      "loss": 0.2351,
+      "step": 1835
+    },
+    {
+      "epoch": 2.4157894736842107,
+      "grad_norm": 0.08449582717174542,
+      "learning_rate": 2.222762763569862e-05,
+      "loss": 0.2331,
+      "step": 1836
+    },
+    {
+      "epoch": 2.417105263157895,
+      "grad_norm": 0.08533743090482276,
+      "learning_rate": 2.2131479796594767e-05,
+      "loss": 0.2356,
+      "step": 1837
+    },
+    {
+      "epoch": 2.418421052631579,
+      "grad_norm": 0.08874843333228837,
+      "learning_rate": 2.2035514475981756e-05,
+      "loss": 0.242,
+      "step": 1838
+    },
+    {
+      "epoch": 2.419736842105263,
+      "grad_norm": 0.08529644318540644,
+      "learning_rate": 2.1939731898795802e-05,
+      "loss": 0.2239,
+      "step": 1839
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.08594783838232867,
+      "learning_rate": 2.184413228954468e-05,
+      "loss": 0.2411,
+      "step": 1840
+    },
+    {
+      "epoch": 2.4223684210526315,
+      "grad_norm": 0.09034497841922952,
+      "learning_rate": 2.1748715872307345e-05,
+      "loss": 0.2437,
+      "step": 1841
+    },
+    {
+      "epoch": 2.4236842105263157,
+      "grad_norm": 0.0864543032403873,
+      "learning_rate": 2.165348287073339e-05,
+      "loss": 0.2497,
+      "step": 1842
+    },
+    {
+      "epoch": 2.425,
+      "grad_norm": 0.08423651720196439,
+      "learning_rate": 2.155843350804243e-05,
+      "loss": 0.233,
+      "step": 1843
+    },
+    {
+      "epoch": 2.4263157894736844,
+      "grad_norm": 0.08476088806860947,
+      "learning_rate": 2.1463568007023704e-05,
+      "loss": 0.2338,
+      "step": 1844
+    },
+    {
+      "epoch": 2.4276315789473686,
+      "grad_norm": 0.08560695927731492,
+      "learning_rate": 2.1368886590035443e-05,
+      "loss": 0.2474,
+      "step": 1845
+    },
+    {
+      "epoch": 2.4289473684210527,
+      "grad_norm": 0.08571656417086615,
+      "learning_rate": 2.1274389479004397e-05,
+      "loss": 0.2343,
+      "step": 1846
+    },
+    {
+      "epoch": 2.430263157894737,
+      "grad_norm": 0.08484405706280464,
+      "learning_rate": 2.1180076895425395e-05,
+      "loss": 0.2358,
+      "step": 1847
+    },
+    {
+      "epoch": 2.431578947368421,
+      "grad_norm": 0.09045256351090854,
+      "learning_rate": 2.1085949060360654e-05,
+      "loss": 0.241,
+      "step": 1848
+    },
+    {
+      "epoch": 2.432894736842105,
+      "grad_norm": 0.08339645141686791,
+      "learning_rate": 2.0992006194439372e-05,
+      "loss": 0.2417,
+      "step": 1849
+    },
+    {
+      "epoch": 2.4342105263157894,
+      "grad_norm": 0.08857069700987559,
+      "learning_rate": 2.0898248517857256e-05,
+      "loss": 0.2281,
+      "step": 1850
+    },
+    {
+      "epoch": 2.4355263157894735,
+      "grad_norm": 0.08727223363474927,
+      "learning_rate": 2.0804676250375867e-05,
+      "loss": 0.2408,
+      "step": 1851
+    },
+    {
+      "epoch": 2.4368421052631577,
+      "grad_norm": 0.08636083640113454,
+      "learning_rate": 2.0711289611322204e-05,
+      "loss": 0.2357,
+      "step": 1852
+    },
+    {
+      "epoch": 2.4381578947368423,
+      "grad_norm": 0.08601921483257845,
+      "learning_rate": 2.0618088819588167e-05,
+      "loss": 0.2362,
+      "step": 1853
+    },
+    {
+      "epoch": 2.4394736842105265,
+      "grad_norm": 0.08573635555006953,
+      "learning_rate": 2.0525074093630036e-05,
+      "loss": 0.2382,
+      "step": 1854
+    },
+    {
+      "epoch": 2.4407894736842106,
+      "grad_norm": 0.08431162925824574,
+      "learning_rate": 2.0432245651467995e-05,
+      "loss": 0.2336,
+      "step": 1855
+    },
+    {
+      "epoch": 2.442105263157895,
+      "grad_norm": 0.08468414851359736,
+      "learning_rate": 2.033960371068557e-05,
+      "loss": 0.2217,
+      "step": 1856
+    },
+    {
+      "epoch": 2.443421052631579,
+      "grad_norm": 0.08472832758486493,
+      "learning_rate": 2.02471484884291e-05,
+      "loss": 0.2378,
+      "step": 1857
+    },
+    {
+      "epoch": 2.444736842105263,
+      "grad_norm": 0.08936877382897962,
+      "learning_rate": 2.0154880201407367e-05,
+      "loss": 0.2449,
+      "step": 1858
+    },
+    {
+      "epoch": 2.4460526315789473,
+      "grad_norm": 0.0841329121430106,
+      "learning_rate": 2.0062799065890904e-05,
+      "loss": 0.2318,
+      "step": 1859
+    },
+    {
+      "epoch": 2.4473684210526314,
+      "grad_norm": 0.08665657400105448,
+      "learning_rate": 1.9970905297711606e-05,
+      "loss": 0.2404,
+      "step": 1860
+    },
+    {
+      "epoch": 2.4486842105263156,
+      "grad_norm": 0.08891275836559771,
+      "learning_rate": 1.987919911226217e-05,
+      "loss": 0.237,
+      "step": 1861
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.08586405744449396,
+      "learning_rate": 1.9787680724495617e-05,
+      "loss": 0.2316,
+      "step": 1862
+    },
+    {
+      "epoch": 2.4513157894736843,
+      "grad_norm": 0.08602563906517424,
+      "learning_rate": 1.969635034892485e-05,
+      "loss": 0.2357,
+      "step": 1863
+    },
+    {
+      "epoch": 2.4526315789473685,
+      "grad_norm": 0.08605411678673527,
+      "learning_rate": 1.9605208199621995e-05,
+      "loss": 0.2368,
+      "step": 1864
+    },
+    {
+      "epoch": 2.4539473684210527,
+      "grad_norm": 0.0847811039119867,
+      "learning_rate": 1.9514254490218e-05,
+      "loss": 0.2278,
+      "step": 1865
+    },
+    {
+      "epoch": 2.455263157894737,
+      "grad_norm": 0.089108249737477,
+      "learning_rate": 1.9423489433902186e-05,
+      "loss": 0.2421,
+      "step": 1866
+    },
+    {
+      "epoch": 2.456578947368421,
+      "grad_norm": 0.08853862535589128,
+      "learning_rate": 1.9332913243421634e-05,
+      "loss": 0.2346,
+      "step": 1867
+    },
+    {
+      "epoch": 2.457894736842105,
+      "grad_norm": 0.08624747222459084,
+      "learning_rate": 1.924252613108073e-05,
+      "loss": 0.2385,
+      "step": 1868
+    },
+    {
+      "epoch": 2.4592105263157893,
+      "grad_norm": 0.08633100421050345,
+      "learning_rate": 1.9152328308740707e-05,
+      "loss": 0.2281,
+      "step": 1869
+    },
+    {
+      "epoch": 2.4605263157894735,
+      "grad_norm": 0.0823894382169209,
+      "learning_rate": 1.9062319987819067e-05,
+      "loss": 0.2313,
+      "step": 1870
+    },
+    {
+      "epoch": 2.461842105263158,
+      "grad_norm": 0.08514975111544096,
+      "learning_rate": 1.897250137928921e-05,
+      "loss": 0.2301,
+      "step": 1871
+    },
+    {
+      "epoch": 2.463157894736842,
+      "grad_norm": 0.08573391873752553,
+      "learning_rate": 1.888287269367979e-05,
+      "loss": 0.2253,
+      "step": 1872
+    },
+    {
+      "epoch": 2.4644736842105264,
+      "grad_norm": 0.08763954109623806,
+      "learning_rate": 1.8793434141074295e-05,
+      "loss": 0.2446,
+      "step": 1873
+    },
+    {
+      "epoch": 2.4657894736842105,
+      "grad_norm": 0.0848193021797019,
+      "learning_rate": 1.870418593111064e-05,
+      "loss": 0.2421,
+      "step": 1874
+    },
+    {
+      "epoch": 2.4671052631578947,
+      "grad_norm": 0.08299936916001775,
+      "learning_rate": 1.861512827298051e-05,
+      "loss": 0.2377,
+      "step": 1875
+    },
+    {
+      "epoch": 2.468421052631579,
+      "grad_norm": 0.08656696401202549,
+      "learning_rate": 1.8526261375428955e-05,
+      "loss": 0.2453,
+      "step": 1876
+    },
+    {
+      "epoch": 2.469736842105263,
+      "grad_norm": 0.0845516003224581,
+      "learning_rate": 1.8437585446753925e-05,
+      "loss": 0.2264,
+      "step": 1877
+    },
+    {
+      "epoch": 2.4710526315789476,
+      "grad_norm": 0.08820431143873694,
+      "learning_rate": 1.834910069480571e-05,
+      "loss": 0.2553,
+      "step": 1878
+    },
+    {
+      "epoch": 2.4723684210526318,
+      "grad_norm": 0.086680366075887,
+      "learning_rate": 1.826080732698656e-05,
+      "loss": 0.2332,
+      "step": 1879
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.08757999419788688,
+      "learning_rate": 1.8172705550250092e-05,
+      "loss": 0.2365,
+      "step": 1880
+    },
+    {
+      "epoch": 2.475,
+      "grad_norm": 0.0819273273150161,
+      "learning_rate": 1.808479557110081e-05,
+      "loss": 0.2222,
+      "step": 1881
+    },
+    {
+      "epoch": 2.4763157894736842,
+      "grad_norm": 0.08333326519630053,
+      "learning_rate": 1.799707759559376e-05,
+      "loss": 0.2393,
+      "step": 1882
+    },
+    {
+      "epoch": 2.4776315789473684,
+      "grad_norm": 0.08611320760984689,
+      "learning_rate": 1.790955182933385e-05,
+      "loss": 0.2268,
+      "step": 1883
+    },
+    {
+      "epoch": 2.4789473684210526,
+      "grad_norm": 0.08633530140680602,
+      "learning_rate": 1.7822218477475494e-05,
+      "loss": 0.236,
+      "step": 1884
+    },
+    {
+      "epoch": 2.4802631578947367,
+      "grad_norm": 0.08652135837244138,
+      "learning_rate": 1.7735077744722107e-05,
+      "loss": 0.2351,
+      "step": 1885
+    },
+    {
+      "epoch": 2.481578947368421,
+      "grad_norm": 0.08815435855670466,
+      "learning_rate": 1.7648129835325587e-05,
+      "loss": 0.2347,
+      "step": 1886
+    },
+    {
+      "epoch": 2.4828947368421055,
+      "grad_norm": 0.0879883933592944,
+      "learning_rate": 1.756137495308594e-05,
+      "loss": 0.2328,
+      "step": 1887
+    },
+    {
+      "epoch": 2.4842105263157896,
+      "grad_norm": 0.08643787173876677,
+      "learning_rate": 1.7474813301350666e-05,
+      "loss": 0.2387,
+      "step": 1888
+    },
+    {
+      "epoch": 2.485526315789474,
+      "grad_norm": 0.08893052591767589,
+      "learning_rate": 1.7388445083014325e-05,
+      "loss": 0.2428,
+      "step": 1889
+    },
+    {
+      "epoch": 2.486842105263158,
+      "grad_norm": 0.08536802381106222,
+      "learning_rate": 1.7302270500518182e-05,
+      "loss": 0.2375,
+      "step": 1890
+    },
+    {
+      "epoch": 2.488157894736842,
+      "grad_norm": 0.08703152842078288,
+      "learning_rate": 1.7216289755849525e-05,
+      "loss": 0.2308,
+      "step": 1891
+    },
+    {
+      "epoch": 2.4894736842105263,
+      "grad_norm": 0.0836327066485613,
+      "learning_rate": 1.7130503050541368e-05,
+      "loss": 0.2363,
+      "step": 1892
+    },
+    {
+      "epoch": 2.4907894736842104,
+      "grad_norm": 0.08521480449342837,
+      "learning_rate": 1.704491058567187e-05,
+      "loss": 0.2371,
+      "step": 1893
+    },
+    {
+      "epoch": 2.4921052631578946,
+      "grad_norm": 0.08544333339590796,
+      "learning_rate": 1.6959512561863912e-05,
+      "loss": 0.2376,
+      "step": 1894
+    },
+    {
+      "epoch": 2.4934210526315788,
+      "grad_norm": 0.08419989528831981,
+      "learning_rate": 1.6874309179284664e-05,
+      "loss": 0.233,
+      "step": 1895
+    },
+    {
+      "epoch": 2.4947368421052634,
+      "grad_norm": 0.08743178517536454,
+      "learning_rate": 1.6789300637645e-05,
+      "loss": 0.24,
+      "step": 1896
+    },
+    {
+      "epoch": 2.4960526315789475,
+      "grad_norm": 0.08627446891742664,
+      "learning_rate": 1.670448713619913e-05,
+      "loss": 0.2418,
+      "step": 1897
+    },
+    {
+      "epoch": 2.4973684210526317,
+      "grad_norm": 0.086520690921376,
+      "learning_rate": 1.6619868873744147e-05,
+      "loss": 0.2333,
+      "step": 1898
+    },
+    {
+      "epoch": 2.498684210526316,
+      "grad_norm": 0.0890053931742651,
+      "learning_rate": 1.653544604861945e-05,
+      "loss": 0.245,
+      "step": 1899
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.08795310211457955,
+      "learning_rate": 1.6451218858706374e-05,
+      "loss": 0.2424,
+      "step": 1900
+    },
+    {
+      "epoch": 2.501315789473684,
+      "grad_norm": 0.08768860420319245,
+      "learning_rate": 1.6367187501427685e-05,
+      "loss": 0.2409,
+      "step": 1901
+    },
+    {
+      "epoch": 2.5026315789473683,
+      "grad_norm": 0.08433281148019468,
+      "learning_rate": 1.6283352173747145e-05,
+      "loss": 0.2332,
+      "step": 1902
+    },
+    {
+      "epoch": 2.5039473684210525,
+      "grad_norm": 0.08869485067176774,
+      "learning_rate": 1.6199713072169053e-05,
+      "loss": 0.247,
+      "step": 1903
+    },
+    {
+      "epoch": 2.5052631578947366,
+      "grad_norm": 0.08535593567657823,
+      "learning_rate": 1.6116270392737754e-05,
+      "loss": 0.2331,
+      "step": 1904
+    },
+    {
+      "epoch": 2.5065789473684212,
+      "grad_norm": 0.08375689041638204,
+      "learning_rate": 1.6033024331037138e-05,
+      "loss": 0.2275,
+      "step": 1905
+    },
+    {
+      "epoch": 2.5078947368421054,
+      "grad_norm": 0.08478044933594917,
+      "learning_rate": 1.5949975082190337e-05,
+      "loss": 0.2339,
+      "step": 1906
+    },
+    {
+      "epoch": 2.5092105263157896,
+      "grad_norm": 0.08731008917029157,
+      "learning_rate": 1.5867122840859117e-05,
+      "loss": 0.2345,
+      "step": 1907
+    },
+    {
+      "epoch": 2.5105263157894737,
+      "grad_norm": 0.0867834465835224,
+      "learning_rate": 1.578446780124344e-05,
+      "loss": 0.2306,
+      "step": 1908
+    },
+    {
+      "epoch": 2.511842105263158,
+      "grad_norm": 0.08484337640412591,
+      "learning_rate": 1.570201015708108e-05,
+      "loss": 0.2421,
+      "step": 1909
+    },
+    {
+      "epoch": 2.513157894736842,
+      "grad_norm": 0.08676705310204419,
+      "learning_rate": 1.5619750101647114e-05,
+      "loss": 0.2431,
+      "step": 1910
+    },
+    {
+      "epoch": 2.514473684210526,
+      "grad_norm": 0.08560817073143388,
+      "learning_rate": 1.553768782775351e-05,
+      "loss": 0.2333,
+      "step": 1911
+    },
+    {
+      "epoch": 2.515789473684211,
+      "grad_norm": 0.08577547460347472,
+      "learning_rate": 1.5455823527748626e-05,
+      "loss": 0.2372,
+      "step": 1912
+    },
+    {
+      "epoch": 2.5171052631578945,
+      "grad_norm": 0.08601962847067497,
+      "learning_rate": 1.5374157393516764e-05,
+      "loss": 0.2334,
+      "step": 1913
+    },
+    {
+      "epoch": 2.518421052631579,
+      "grad_norm": 0.08618206990680746,
+      "learning_rate": 1.5292689616477806e-05,
+      "loss": 0.2355,
+      "step": 1914
+    },
+    {
+      "epoch": 2.5197368421052633,
+      "grad_norm": 0.08561559936263621,
+      "learning_rate": 1.5211420387586638e-05,
+      "loss": 0.2397,
+      "step": 1915
+    },
+    {
+      "epoch": 2.5210526315789474,
+      "grad_norm": 0.08569149401048583,
+      "learning_rate": 1.5130349897332763e-05,
+      "loss": 0.2385,
+      "step": 1916
+    },
+    {
+      "epoch": 2.5223684210526316,
+      "grad_norm": 0.08528987545012465,
+      "learning_rate": 1.5049478335739886e-05,
+      "loss": 0.2422,
+      "step": 1917
+    },
+    {
+      "epoch": 2.5236842105263158,
+      "grad_norm": 0.08573039796052453,
+      "learning_rate": 1.49688058923654e-05,
+      "loss": 0.2346,
+      "step": 1918
+    },
+    {
+      "epoch": 2.525,
+      "grad_norm": 0.08483858064921189,
+      "learning_rate": 1.4888332756300027e-05,
+      "loss": 0.2335,
+      "step": 1919
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.08618837736943744,
+      "learning_rate": 1.4808059116167305e-05,
+      "loss": 0.2321,
+      "step": 1920
+    },
+    {
+      "epoch": 2.5276315789473687,
+      "grad_norm": 0.08306460927365213,
+      "learning_rate": 1.4727985160123114e-05,
+      "loss": 0.2322,
+      "step": 1921
+    },
+    {
+      "epoch": 2.5289473684210524,
+      "grad_norm": 0.08847498270917728,
+      "learning_rate": 1.4648111075855398e-05,
+      "loss": 0.238,
+      "step": 1922
+    },
+    {
+      "epoch": 2.530263157894737,
+      "grad_norm": 0.08664647691974588,
+      "learning_rate": 1.4568437050583517e-05,
+      "loss": 0.2329,
+      "step": 1923
+    },
+    {
+      "epoch": 2.531578947368421,
+      "grad_norm": 0.08925646785401711,
+      "learning_rate": 1.4488963271057943e-05,
+      "loss": 0.2447,
+      "step": 1924
+    },
+    {
+      "epoch": 2.5328947368421053,
+      "grad_norm": 0.0858422178404279,
+      "learning_rate": 1.44096899235598e-05,
+      "loss": 0.2339,
+      "step": 1925
+    },
+    {
+      "epoch": 2.5342105263157895,
+      "grad_norm": 0.085698190947571,
+      "learning_rate": 1.4330617193900364e-05,
+      "loss": 0.2329,
+      "step": 1926
+    },
+    {
+      "epoch": 2.5355263157894736,
+      "grad_norm": 0.08849510463317915,
+      "learning_rate": 1.4251745267420757e-05,
+      "loss": 0.236,
+      "step": 1927
+    },
+    {
+      "epoch": 2.536842105263158,
+      "grad_norm": 0.08642342384169846,
+      "learning_rate": 1.4173074328991377e-05,
+      "loss": 0.2292,
+      "step": 1928
+    },
+    {
+      "epoch": 2.538157894736842,
+      "grad_norm": 0.08593496440244164,
+      "learning_rate": 1.4094604563011472e-05,
+      "loss": 0.2308,
+      "step": 1929
+    },
+    {
+      "epoch": 2.5394736842105265,
+      "grad_norm": 0.08537576961097464,
+      "learning_rate": 1.4016336153408893e-05,
+      "loss": 0.2318,
+      "step": 1930
+    },
+    {
+      "epoch": 2.5407894736842103,
+      "grad_norm": 0.08435630547007106,
+      "learning_rate": 1.3938269283639394e-05,
+      "loss": 0.2244,
+      "step": 1931
+    },
+    {
+      "epoch": 2.542105263157895,
+      "grad_norm": 0.0894275649268176,
+      "learning_rate": 1.3860404136686411e-05,
+      "loss": 0.2386,
+      "step": 1932
+    },
+    {
+      "epoch": 2.543421052631579,
+      "grad_norm": 0.08731027043438779,
+      "learning_rate": 1.3782740895060497e-05,
+      "loss": 0.2438,
+      "step": 1933
+    },
+    {
+      "epoch": 2.544736842105263,
+      "grad_norm": 0.08677749818948599,
+      "learning_rate": 1.3705279740798993e-05,
+      "loss": 0.2383,
+      "step": 1934
+    },
+    {
+      "epoch": 2.5460526315789473,
+      "grad_norm": 0.08591491161614231,
+      "learning_rate": 1.3628020855465572e-05,
+      "loss": 0.2387,
+      "step": 1935
+    },
+    {
+      "epoch": 2.5473684210526315,
+      "grad_norm": 0.08575789115259161,
+      "learning_rate": 1.355096442014977e-05,
+      "loss": 0.2325,
+      "step": 1936
+    },
+    {
+      "epoch": 2.5486842105263157,
+      "grad_norm": 0.08651010282289677,
+      "learning_rate": 1.3474110615466583e-05,
+      "loss": 0.2279,
+      "step": 1937
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.08869408109954365,
+      "learning_rate": 1.339745962155613e-05,
+      "loss": 0.2351,
+      "step": 1938
+    },
+    {
+      "epoch": 2.5513157894736844,
+      "grad_norm": 0.08792931672329181,
+      "learning_rate": 1.3321011618083079e-05,
+      "loss": 0.2439,
+      "step": 1939
+    },
+    {
+      "epoch": 2.5526315789473686,
+      "grad_norm": 0.08816455279822999,
+      "learning_rate": 1.3244766784236307e-05,
+      "loss": 0.2429,
+      "step": 1940
+    },
+    {
+      "epoch": 2.5539473684210527,
+      "grad_norm": 0.08652728797542152,
+      "learning_rate": 1.3168725298728524e-05,
+      "loss": 0.2322,
+      "step": 1941
+    },
+    {
+      "epoch": 2.555263157894737,
+      "grad_norm": 0.08646822510404321,
+      "learning_rate": 1.3092887339795734e-05,
+      "loss": 0.2448,
+      "step": 1942
+    },
+    {
+      "epoch": 2.556578947368421,
+      "grad_norm": 0.08827748125160723,
+      "learning_rate": 1.3017253085196979e-05,
+      "loss": 0.2432,
+      "step": 1943
+    },
+    {
+      "epoch": 2.557894736842105,
+      "grad_norm": 0.087400695177381,
+      "learning_rate": 1.294182271221377e-05,
+      "loss": 0.242,
+      "step": 1944
+    },
+    {
+      "epoch": 2.5592105263157894,
+      "grad_norm": 0.08476067656041161,
+      "learning_rate": 1.2866596397649721e-05,
+      "loss": 0.2332,
+      "step": 1945
+    },
+    {
+      "epoch": 2.5605263157894735,
+      "grad_norm": 0.08400036515218953,
+      "learning_rate": 1.2791574317830213e-05,
+      "loss": 0.2335,
+      "step": 1946
+    },
+    {
+      "epoch": 2.5618421052631577,
+      "grad_norm": 0.08714320110806623,
+      "learning_rate": 1.2716756648601857e-05,
+      "loss": 0.2384,
+      "step": 1947
+    },
+    {
+      "epoch": 2.5631578947368423,
+      "grad_norm": 0.08718032193532914,
+      "learning_rate": 1.2642143565332154e-05,
+      "loss": 0.2363,
+      "step": 1948
+    },
+    {
+      "epoch": 2.5644736842105265,
+      "grad_norm": 0.08459275476172294,
+      "learning_rate": 1.2567735242909074e-05,
+      "loss": 0.2226,
+      "step": 1949
+    },
+    {
+      "epoch": 2.5657894736842106,
+      "grad_norm": 0.08586849969458207,
+      "learning_rate": 1.2493531855740625e-05,
+      "loss": 0.2247,
+      "step": 1950
+    },
+    {
+      "epoch": 2.567105263157895,
+      "grad_norm": 0.08846213083149775,
+      "learning_rate": 1.2419533577754528e-05,
+      "loss": 0.2333,
+      "step": 1951
+    },
+    {
+      "epoch": 2.568421052631579,
+      "grad_norm": 0.08643152306184454,
+      "learning_rate": 1.2345740582397648e-05,
+      "loss": 0.2371,
+      "step": 1952
+    },
+    {
+      "epoch": 2.569736842105263,
+      "grad_norm": 0.08661338412832961,
+      "learning_rate": 1.2272153042635704e-05,
+      "loss": 0.2434,
+      "step": 1953
+    },
+    {
+      "epoch": 2.5710526315789473,
+      "grad_norm": 0.08897240710261847,
+      "learning_rate": 1.2198771130952913e-05,
+      "loss": 0.2376,
+      "step": 1954
+    },
+    {
+      "epoch": 2.5723684210526314,
+      "grad_norm": 0.08622184838413748,
+      "learning_rate": 1.2125595019351443e-05,
+      "loss": 0.2307,
+      "step": 1955
+    },
+    {
+      "epoch": 2.5736842105263156,
+      "grad_norm": 0.0907063460852498,
+      "learning_rate": 1.2052624879351104e-05,
+      "loss": 0.2313,
+      "step": 1956
+    },
+    {
+      "epoch": 2.575,
+      "grad_norm": 0.08625315493284626,
+      "learning_rate": 1.1979860881988902e-05,
+      "loss": 0.2383,
+      "step": 1957
+    },
+    {
+      "epoch": 2.5763157894736843,
+      "grad_norm": 0.08964244854263959,
+      "learning_rate": 1.1907303197818665e-05,
+      "loss": 0.2394,
+      "step": 1958
+    },
+    {
+      "epoch": 2.5776315789473685,
+      "grad_norm": 0.08330532747614758,
+      "learning_rate": 1.183495199691068e-05,
+      "loss": 0.2215,
+      "step": 1959
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.0884851429344919,
+      "learning_rate": 1.176280744885121e-05,
+      "loss": 0.2423,
+      "step": 1960
+    },
+    {
+      "epoch": 2.580263157894737,
+      "grad_norm": 0.08980407385361433,
+      "learning_rate": 1.1690869722742126e-05,
+      "loss": 0.2393,
+      "step": 1961
+    },
+    {
+      "epoch": 2.581578947368421,
+      "grad_norm": 0.08634666525024143,
+      "learning_rate": 1.1619138987200562e-05,
+      "loss": 0.2287,
+      "step": 1962
+    },
+    {
+      "epoch": 2.582894736842105,
+      "grad_norm": 0.08589844097584014,
+      "learning_rate": 1.154761541035847e-05,
+      "loss": 0.2353,
+      "step": 1963
+    },
+    {
+      "epoch": 2.5842105263157897,
+      "grad_norm": 0.08695619199252902,
+      "learning_rate": 1.1476299159862203e-05,
+      "loss": 0.2424,
+      "step": 1964
+    },
+    {
+      "epoch": 2.5855263157894735,
+      "grad_norm": 0.08435657255170038,
+      "learning_rate": 1.1405190402872202e-05,
+      "loss": 0.2308,
+      "step": 1965
+    },
+    {
+      "epoch": 2.586842105263158,
+      "grad_norm": 0.08681956736459753,
+      "learning_rate": 1.1334289306062495e-05,
+      "loss": 0.2387,
+      "step": 1966
+    },
+    {
+      "epoch": 2.588157894736842,
+      "grad_norm": 0.08650224803441879,
+      "learning_rate": 1.126359603562045e-05,
+      "loss": 0.2369,
+      "step": 1967
+    },
+    {
+      "epoch": 2.5894736842105264,
+      "grad_norm": 0.08441575091576375,
+      "learning_rate": 1.119311075724625e-05,
+      "loss": 0.2195,
+      "step": 1968
+    },
+    {
+      "epoch": 2.5907894736842105,
+      "grad_norm": 0.08371697252642783,
+      "learning_rate": 1.1122833636152563e-05,
+      "loss": 0.2174,
+      "step": 1969
+    },
+    {
+      "epoch": 2.5921052631578947,
+      "grad_norm": 0.0868737302053953,
+      "learning_rate": 1.1052764837064178e-05,
+      "loss": 0.2283,
+      "step": 1970
+    },
+    {
+      "epoch": 2.593421052631579,
+      "grad_norm": 0.08687526871015291,
+      "learning_rate": 1.0982904524217551e-05,
+      "loss": 0.2291,
+      "step": 1971
+    },
+    {
+      "epoch": 2.594736842105263,
+      "grad_norm": 0.08813319422928485,
+      "learning_rate": 1.09132528613605e-05,
+      "loss": 0.2449,
+      "step": 1972
+    },
+    {
+      "epoch": 2.5960526315789476,
+      "grad_norm": 0.08856476761244737,
+      "learning_rate": 1.0843810011751766e-05,
+      "loss": 0.2422,
+      "step": 1973
+    },
+    {
+      "epoch": 2.5973684210526313,
+      "grad_norm": 0.08769322323788943,
+      "learning_rate": 1.0774576138160597e-05,
+      "loss": 0.245,
+      "step": 1974
+    },
+    {
+      "epoch": 2.598684210526316,
+      "grad_norm": 0.0854386832496123,
+      "learning_rate": 1.070555140286652e-05,
+      "loss": 0.2269,
+      "step": 1975
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.08380866243911338,
+      "learning_rate": 1.0636735967658784e-05,
+      "loss": 0.239,
+      "step": 1976
+    },
+    {
+      "epoch": 2.6013157894736842,
+      "grad_norm": 0.08479138359498807,
+      "learning_rate": 1.056812999383604e-05,
+      "loss": 0.2321,
+      "step": 1977
+    },
+    {
+      "epoch": 2.6026315789473684,
+      "grad_norm": 0.08709294911680168,
+      "learning_rate": 1.0499733642206033e-05,
+      "loss": 0.239,
+      "step": 1978
+    },
+    {
+      "epoch": 2.6039473684210526,
+      "grad_norm": 0.08505697002044242,
+      "learning_rate": 1.0431547073085135e-05,
+      "loss": 0.2314,
+      "step": 1979
+    },
+    {
+      "epoch": 2.6052631578947367,
+      "grad_norm": 0.08770454123258202,
+      "learning_rate": 1.0363570446297999e-05,
+      "loss": 0.2455,
+      "step": 1980
+    },
+    {
+      "epoch": 2.606578947368421,
+      "grad_norm": 0.08477772700319303,
+      "learning_rate": 1.0295803921177182e-05,
+      "loss": 0.2277,
+      "step": 1981
+    },
+    {
+      "epoch": 2.6078947368421055,
+      "grad_norm": 0.08575690009081015,
+      "learning_rate": 1.0228247656562795e-05,
+      "loss": 0.2352,
+      "step": 1982
+    },
+    {
+      "epoch": 2.609210526315789,
+      "grad_norm": 0.0852766947310931,
+      "learning_rate": 1.0160901810802115e-05,
+      "loss": 0.2379,
+      "step": 1983
+    },
+    {
+      "epoch": 2.610526315789474,
+      "grad_norm": 0.08360204366625848,
+      "learning_rate": 1.0093766541749205e-05,
+      "loss": 0.2266,
+      "step": 1984
+    },
+    {
+      "epoch": 2.611842105263158,
+      "grad_norm": 0.08424512848837429,
+      "learning_rate": 1.0026842006764526e-05,
+      "loss": 0.2307,
+      "step": 1985
+    },
+    {
+      "epoch": 2.613157894736842,
+      "grad_norm": 0.08406138879126103,
+      "learning_rate": 9.960128362714637e-06,
+      "loss": 0.2293,
+      "step": 1986
+    },
+    {
+      "epoch": 2.6144736842105263,
+      "grad_norm": 0.08644831073662886,
+      "learning_rate": 9.89362576597177e-06,
+      "loss": 0.2334,
+      "step": 1987
+    },
+    {
+      "epoch": 2.6157894736842104,
+      "grad_norm": 0.08623901530925424,
+      "learning_rate": 9.827334372413444e-06,
+      "loss": 0.2358,
+      "step": 1988
+    },
+    {
+      "epoch": 2.6171052631578946,
+      "grad_norm": 0.08269859922491032,
+      "learning_rate": 9.761254337422176e-06,
+      "loss": 0.2167,
+      "step": 1989
+    },
+    {
+      "epoch": 2.6184210526315788,
+      "grad_norm": 0.08644342262890786,
+      "learning_rate": 9.695385815885016e-06,
+      "loss": 0.2301,
+      "step": 1990
+    },
+    {
+      "epoch": 2.6197368421052634,
+      "grad_norm": 0.08629959299387181,
+      "learning_rate": 9.629728962193318e-06,
+      "loss": 0.2289,
+      "step": 1991
+    },
+    {
+      "epoch": 2.6210526315789475,
+      "grad_norm": 0.08512850244099619,
+      "learning_rate": 9.564283930242257e-06,
+      "loss": 0.2251,
+      "step": 1992
+    },
+    {
+      "epoch": 2.6223684210526317,
+      "grad_norm": 0.0855143103873754,
+      "learning_rate": 9.499050873430482e-06,
+      "loss": 0.2253,
+      "step": 1993
+    },
+    {
+      "epoch": 2.623684210526316,
+      "grad_norm": 0.08812972174025184,
+      "learning_rate": 9.434029944659872e-06,
+      "loss": 0.2227,
+      "step": 1994
+    },
+    {
+      "epoch": 2.625,
+      "grad_norm": 0.08923681105940687,
+      "learning_rate": 9.369221296335006e-06,
+      "loss": 0.2519,
+      "step": 1995
+    },
+    {
+      "epoch": 2.626315789473684,
+      "grad_norm": 0.08888942368903012,
+      "learning_rate": 9.30462508036294e-06,
+      "loss": 0.2451,
+      "step": 1996
+    },
+    {
+      "epoch": 2.6276315789473683,
+      "grad_norm": 0.08691289008570331,
+      "learning_rate": 9.240241448152787e-06,
+      "loss": 0.235,
+      "step": 1997
+    },
+    {
+      "epoch": 2.6289473684210525,
+      "grad_norm": 0.08587149284301165,
+      "learning_rate": 9.176070550615378e-06,
+      "loss": 0.2273,
+      "step": 1998
+    },
+    {
+      "epoch": 2.6302631578947366,
+      "grad_norm": 0.08822548058662702,
+      "learning_rate": 9.112112538162898e-06,
+      "loss": 0.2356,
+      "step": 1999
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.08348460888767743,
+      "learning_rate": 9.048367560708604e-06,
+      "loss": 0.2297,
+      "step": 2000
+    },
+    {
+      "epoch": 2.6328947368421054,
+      "grad_norm": 0.08551510433915722,
+      "learning_rate": 8.98483576766631e-06,
+      "loss": 0.2324,
+      "step": 2001
+    },
+    {
+      "epoch": 2.6342105263157896,
+      "grad_norm": 0.08388165924617051,
+      "learning_rate": 8.921517307950255e-06,
+      "loss": 0.229,
+      "step": 2002
+    },
+    {
+      "epoch": 2.6355263157894737,
+      "grad_norm": 0.08896318131104192,
+      "learning_rate": 8.858412329974552e-06,
+      "loss": 0.2385,
+      "step": 2003
+    },
+    {
+      "epoch": 2.636842105263158,
+      "grad_norm": 0.09015074541095214,
+      "learning_rate": 8.795520981652961e-06,
+      "loss": 0.2392,
+      "step": 2004
+    },
+    {
+      "epoch": 2.638157894736842,
+      "grad_norm": 0.08696268311755063,
+      "learning_rate": 8.732843410398506e-06,
+      "loss": 0.2294,
+      "step": 2005
+    },
+    {
+      "epoch": 2.639473684210526,
+      "grad_norm": 0.08769627213456496,
+      "learning_rate": 8.670379763123126e-06,
+      "loss": 0.2384,
+      "step": 2006
+    },
+    {
+      "epoch": 2.640789473684211,
+      "grad_norm": 0.08643115429762364,
+      "learning_rate": 8.608130186237329e-06,
+      "loss": 0.25,
+      "step": 2007
+    },
+    {
+      "epoch": 2.6421052631578945,
+      "grad_norm": 0.08746848122982008,
+      "learning_rate": 8.546094825649908e-06,
+      "loss": 0.2398,
+      "step": 2008
+    },
+    {
+      "epoch": 2.643421052631579,
+      "grad_norm": 0.08700832769044184,
+      "learning_rate": 8.484273826767474e-06,
+      "loss": 0.2315,
+      "step": 2009
+    },
+    {
+      "epoch": 2.6447368421052633,
+      "grad_norm": 0.08453786644008071,
+      "learning_rate": 8.422667334494249e-06,
+      "loss": 0.2285,
+      "step": 2010
+    },
+    {
+      "epoch": 2.6460526315789474,
+      "grad_norm": 0.08586615693833513,
+      "learning_rate": 8.361275493231646e-06,
+      "loss": 0.2364,
+      "step": 2011
+    },
+    {
+      "epoch": 2.6473684210526316,
+      "grad_norm": 0.0870712891159542,
+      "learning_rate": 8.300098446877923e-06,
+      "loss": 0.2265,
+      "step": 2012
+    },
+    {
+      "epoch": 2.6486842105263158,
+      "grad_norm": 0.08440192942964624,
+      "learning_rate": 8.239136338827903e-06,
+      "loss": 0.2291,
+      "step": 2013
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.08705307636244689,
+      "learning_rate": 8.178389311972612e-06,
+      "loss": 0.2392,
+      "step": 2014
+    },
+    {
+      "epoch": 2.651315789473684,
+      "grad_norm": 0.08894269433728565,
+      "learning_rate": 8.1178575086989e-06,
+      "loss": 0.2315,
+      "step": 2015
+    },
+    {
+      "epoch": 2.6526315789473687,
+      "grad_norm": 0.08919405616773922,
+      "learning_rate": 8.05754107088923e-06,
+      "loss": 0.2429,
+      "step": 2016
+    },
+    {
+      "epoch": 2.6539473684210524,
+      "grad_norm": 0.08652861044002905,
+      "learning_rate": 7.997440139921152e-06,
+      "loss": 0.2386,
+      "step": 2017
+    },
+    {
+      "epoch": 2.655263157894737,
+      "grad_norm": 0.08615918194801953,
+      "learning_rate": 7.937554856667196e-06,
+      "loss": 0.2275,
+      "step": 2018
+    },
+    {
+      "epoch": 2.656578947368421,
+      "grad_norm": 0.0867406788792665,
+      "learning_rate": 7.877885361494353e-06,
+      "loss": 0.2294,
+      "step": 2019
+    },
+    {
+      "epoch": 2.6578947368421053,
+      "grad_norm": 0.08726072526321839,
+      "learning_rate": 7.818431794263836e-06,
+      "loss": 0.236,
+      "step": 2020
+    },
+    {
+      "epoch": 2.6592105263157895,
+      "grad_norm": 0.08547170774404624,
+      "learning_rate": 7.759194294330751e-06,
+      "loss": 0.2416,
+      "step": 2021
+    },
+    {
+      "epoch": 2.6605263157894736,
+      "grad_norm": 0.08743733901669834,
+      "learning_rate": 7.700173000543742e-06,
+      "loss": 0.2409,
+      "step": 2022
+    },
+    {
+      "epoch": 2.661842105263158,
+      "grad_norm": 0.08978412000343398,
+      "learning_rate": 7.641368051244679e-06,
+      "loss": 0.2303,
+      "step": 2023
+    },
+    {
+      "epoch": 2.663157894736842,
+      "grad_norm": 0.08595515849525466,
+      "learning_rate": 7.582779584268373e-06,
+      "loss": 0.2392,
+      "step": 2024
+    },
+    {
+      "epoch": 2.6644736842105265,
+      "grad_norm": 0.08931169378561088,
+      "learning_rate": 7.524407736942174e-06,
+      "loss": 0.236,
+      "step": 2025
+    },
+    {
+      "epoch": 2.6657894736842103,
+      "grad_norm": 0.08550626450400388,
+      "learning_rate": 7.466252646085703e-06,
+      "loss": 0.2351,
+      "step": 2026
+    },
+    {
+      "epoch": 2.667105263157895,
+      "grad_norm": 0.08787743584655944,
+      "learning_rate": 7.4083144480105335e-06,
+      "loss": 0.2464,
+      "step": 2027
+    },
+    {
+      "epoch": 2.668421052631579,
+      "grad_norm": 0.0875076538218403,
+      "learning_rate": 7.350593278519824e-06,
+      "loss": 0.238,
+      "step": 2028
+    },
+    {
+      "epoch": 2.669736842105263,
+      "grad_norm": 0.09154668653402206,
+      "learning_rate": 7.2930892729080716e-06,
+      "loss": 0.2381,
+      "step": 2029
+    },
+    {
+      "epoch": 2.6710526315789473,
+      "grad_norm": 0.08720851842242237,
+      "learning_rate": 7.235802565960714e-06,
+      "loss": 0.2394,
+      "step": 2030
+    },
+    {
+      "epoch": 2.6723684210526315,
+      "grad_norm": 0.08749018793861221,
+      "learning_rate": 7.178733291953865e-06,
+      "loss": 0.2331,
+      "step": 2031
+    },
+    {
+      "epoch": 2.6736842105263157,
+      "grad_norm": 0.08792096320231124,
+      "learning_rate": 7.121881584654056e-06,
+      "loss": 0.2415,
+      "step": 2032
+    },
+    {
+      "epoch": 2.675,
+      "grad_norm": 0.08722269899883195,
+      "learning_rate": 7.0652475773177464e-06,
+      "loss": 0.236,
+      "step": 2033
+    },
+    {
+      "epoch": 2.6763157894736844,
+      "grad_norm": 0.08769168784492533,
+      "learning_rate": 7.00883140269123e-06,
+      "loss": 0.2409,
+      "step": 2034
+    },
+    {
+      "epoch": 2.6776315789473686,
+      "grad_norm": 0.09065018938662237,
+      "learning_rate": 6.95263319301015e-06,
+      "loss": 0.244,
+      "step": 2035
+    },
+    {
+      "epoch": 2.6789473684210527,
+      "grad_norm": 0.08628478345157022,
+      "learning_rate": 6.896653079999249e-06,
+      "loss": 0.2418,
+      "step": 2036
+    },
+    {
+      "epoch": 2.680263157894737,
+      "grad_norm": 0.0847424030869481,
+      "learning_rate": 6.840891194872112e-06,
+      "loss": 0.2367,
+      "step": 2037
+    },
+    {
+      "epoch": 2.681578947368421,
+      "grad_norm": 0.09082411672610778,
+      "learning_rate": 6.785347668330777e-06,
+      "loss": 0.2496,
+      "step": 2038
+    },
+    {
+      "epoch": 2.682894736842105,
+      "grad_norm": 0.08829134391849804,
+      "learning_rate": 6.730022630565458e-06,
+      "loss": 0.242,
+      "step": 2039
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.08183719511764015,
+      "learning_rate": 6.674916211254289e-06,
+      "loss": 0.2239,
+      "step": 2040
+    },
+    {
+      "epoch": 2.6855263157894735,
+      "grad_norm": 0.08448377337525598,
+      "learning_rate": 6.620028539562939e-06,
+      "loss": 0.2306,
+      "step": 2041
+    },
+    {
+      "epoch": 2.6868421052631577,
+      "grad_norm": 0.0850861974045535,
+      "learning_rate": 6.565359744144373e-06,
+      "loss": 0.2434,
+      "step": 2042
+    },
+    {
+      "epoch": 2.6881578947368423,
+      "grad_norm": 0.08652193833655933,
+      "learning_rate": 6.510909953138511e-06,
+      "loss": 0.2399,
+      "step": 2043
+    },
+    {
+      "epoch": 2.6894736842105265,
+      "grad_norm": 0.08571622313764427,
+      "learning_rate": 6.45667929417193e-06,
+      "loss": 0.2168,
+      "step": 2044
+    },
+    {
+      "epoch": 2.6907894736842106,
+      "grad_norm": 0.0891461015547161,
+      "learning_rate": 6.402667894357595e-06,
+      "loss": 0.2358,
+      "step": 2045
+    },
+    {
+      "epoch": 2.692105263157895,
+      "grad_norm": 0.08690415491550392,
+      "learning_rate": 6.3488758802945354e-06,
+      "loss": 0.2306,
+      "step": 2046
+    },
+    {
+      "epoch": 2.693421052631579,
+      "grad_norm": 0.08712160901508374,
+      "learning_rate": 6.2953033780675406e-06,
+      "loss": 0.2435,
+      "step": 2047
+    },
+    {
+      "epoch": 2.694736842105263,
+      "grad_norm": 0.08600309013612613,
+      "learning_rate": 6.2419505132469305e-06,
+      "loss": 0.2348,
+      "step": 2048
+    },
+    {
+      "epoch": 2.6960526315789473,
+      "grad_norm": 0.08761511265084414,
+      "learning_rate": 6.188817410888148e-06,
+      "loss": 0.2362,
+      "step": 2049
+    },
+    {
+      "epoch": 2.6973684210526314,
+      "grad_norm": 0.08786468959375783,
+      "learning_rate": 6.1359041955315725e-06,
+      "loss": 0.228,
+      "step": 2050
+    },
+    {
+      "epoch": 2.6986842105263156,
+      "grad_norm": 0.08424404719132456,
+      "learning_rate": 6.083210991202148e-06,
+      "loss": 0.2309,
+      "step": 2051
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.08855984411695608,
+      "learning_rate": 6.030737921409169e-06,
+      "loss": 0.2481,
+      "step": 2052
+    },
+    {
+      "epoch": 2.7013157894736843,
+      "grad_norm": 0.08622212781661236,
+      "learning_rate": 5.978485109145904e-06,
+      "loss": 0.2316,
+      "step": 2053
+    },
+    {
+      "epoch": 2.7026315789473685,
+      "grad_norm": 0.0877254168652046,
+      "learning_rate": 5.926452676889383e-06,
+      "loss": 0.2442,
+      "step": 2054
+    },
+    {
+      "epoch": 2.7039473684210527,
+      "grad_norm": 0.08959699248737611,
+      "learning_rate": 5.8746407466000464e-06,
+      "loss": 0.2334,
+      "step": 2055
+    },
+    {
+      "epoch": 2.705263157894737,
+      "grad_norm": 0.08448005747326008,
+      "learning_rate": 5.823049439721561e-06,
+      "loss": 0.2246,
+      "step": 2056
+    },
+    {
+      "epoch": 2.706578947368421,
+      "grad_norm": 0.087622602007786,
+      "learning_rate": 5.771678877180408e-06,
+      "loss": 0.2301,
+      "step": 2057
+    },
+    {
+      "epoch": 2.707894736842105,
+      "grad_norm": 0.08772663242966719,
+      "learning_rate": 5.720529179385659e-06,
+      "loss": 0.2335,
+      "step": 2058
+    },
+    {
+      "epoch": 2.7092105263157897,
+      "grad_norm": 0.08569105342372876,
+      "learning_rate": 5.669600466228742e-06,
+      "loss": 0.2334,
+      "step": 2059
+    },
+    {
+      "epoch": 2.7105263157894735,
+      "grad_norm": 0.08468267671093112,
+      "learning_rate": 5.618892857083069e-06,
+      "loss": 0.2261,
+      "step": 2060
+    },
+    {
+      "epoch": 2.711842105263158,
+      "grad_norm": 0.08830354382676614,
+      "learning_rate": 5.568406470803799e-06,
+      "loss": 0.2395,
+      "step": 2061
+    },
+    {
+      "epoch": 2.713157894736842,
+      "grad_norm": 0.08578516507839022,
+      "learning_rate": 5.518141425727586e-06,
+      "loss": 0.228,
+      "step": 2062
+    },
+    {
+      "epoch": 2.7144736842105264,
+      "grad_norm": 0.08452732187319882,
+      "learning_rate": 5.468097839672237e-06,
+      "loss": 0.2139,
+      "step": 2063
+    },
+    {
+      "epoch": 2.7157894736842105,
+      "grad_norm": 0.08925179167431409,
+      "learning_rate": 5.418275829936537e-06,
+      "loss": 0.2473,
+      "step": 2064
+    },
+    {
+      "epoch": 2.7171052631578947,
+      "grad_norm": 0.08802009007215943,
+      "learning_rate": 5.3686755132998475e-06,
+      "loss": 0.2354,
+      "step": 2065
+    },
+    {
+      "epoch": 2.718421052631579,
+      "grad_norm": 0.08638926510806848,
+      "learning_rate": 5.319297006021917e-06,
+      "loss": 0.2276,
+      "step": 2066
+    },
+    {
+      "epoch": 2.719736842105263,
+      "grad_norm": 0.08649040265863565,
+      "learning_rate": 5.270140423842607e-06,
+      "loss": 0.2319,
+      "step": 2067
+    },
+    {
+      "epoch": 2.7210526315789476,
+      "grad_norm": 0.09027369123729026,
+      "learning_rate": 5.221205881981595e-06,
+      "loss": 0.2362,
+      "step": 2068
+    },
+    {
+      "epoch": 2.7223684210526313,
+      "grad_norm": 0.09174442002171228,
+      "learning_rate": 5.1724934951380755e-06,
+      "loss": 0.241,
+      "step": 2069
+    },
+    {
+      "epoch": 2.723684210526316,
+      "grad_norm": 0.08685169046569569,
+      "learning_rate": 5.124003377490582e-06,
+      "loss": 0.2389,
+      "step": 2070
+    },
+    {
+      "epoch": 2.725,
+      "grad_norm": 0.08724300958525942,
+      "learning_rate": 5.075735642696611e-06,
+      "loss": 0.2284,
+      "step": 2071
+    },
+    {
+      "epoch": 2.7263157894736842,
+      "grad_norm": 0.08498489763962802,
+      "learning_rate": 5.02769040389246e-06,
+      "loss": 0.2242,
+      "step": 2072
+    },
+    {
+      "epoch": 2.7276315789473684,
+      "grad_norm": 0.08637681029702944,
+      "learning_rate": 4.979867773692881e-06,
+      "loss": 0.2304,
+      "step": 2073
+    },
+    {
+      "epoch": 2.7289473684210526,
+      "grad_norm": 0.08534225285490787,
+      "learning_rate": 4.932267864190832e-06,
+      "loss": 0.2241,
+      "step": 2074
+    },
+    {
+      "epoch": 2.7302631578947367,
+      "grad_norm": 0.08306302707156428,
+      "learning_rate": 4.884890786957264e-06,
+      "loss": 0.2215,
+      "step": 2075
+    },
+    {
+      "epoch": 2.731578947368421,
+      "grad_norm": 0.08910155948565507,
+      "learning_rate": 4.8377366530408254e-06,
+      "loss": 0.2356,
+      "step": 2076
+    },
+    {
+      "epoch": 2.7328947368421055,
+      "grad_norm": 0.0864142099489445,
+      "learning_rate": 4.790805572967549e-06,
+      "loss": 0.2376,
+      "step": 2077
+    },
+    {
+      "epoch": 2.734210526315789,
+      "grad_norm": 0.08777769993987874,
+      "learning_rate": 4.744097656740709e-06,
+      "loss": 0.2296,
+      "step": 2078
+    },
+    {
+      "epoch": 2.735526315789474,
+      "grad_norm": 0.08927992176441508,
+      "learning_rate": 4.697613013840441e-06,
+      "loss": 0.245,
+      "step": 2079
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.08952627138575911,
+      "learning_rate": 4.65135175322361e-06,
+      "loss": 0.2361,
+      "step": 2080
+    },
+    {
+      "epoch": 2.738157894736842,
+      "grad_norm": 0.0887680485842762,
+      "learning_rate": 4.605313983323423e-06,
+      "loss": 0.2384,
+      "step": 2081
+    },
+    {
+      "epoch": 2.7394736842105263,
+      "grad_norm": 0.08612832481286653,
+      "learning_rate": 4.559499812049251e-06,
+      "loss": 0.2299,
+      "step": 2082
+    },
+    {
+      "epoch": 2.7407894736842104,
+      "grad_norm": 0.08792908730829178,
+      "learning_rate": 4.513909346786427e-06,
+      "loss": 0.24,
+      "step": 2083
+    },
+    {
+      "epoch": 2.7421052631578946,
+      "grad_norm": 0.08726287355056869,
+      "learning_rate": 4.468542694395861e-06,
+      "loss": 0.2355,
+      "step": 2084
+    },
+    {
+      "epoch": 2.7434210526315788,
+      "grad_norm": 0.08377835577746928,
+      "learning_rate": 4.423399961213892e-06,
+      "loss": 0.2242,
+      "step": 2085
+    },
+    {
+      "epoch": 2.7447368421052634,
+      "grad_norm": 0.08736709748947386,
+      "learning_rate": 4.378481253051991e-06,
+      "loss": 0.2313,
+      "step": 2086
+    },
+    {
+      "epoch": 2.7460526315789475,
+      "grad_norm": 0.08840579315321198,
+      "learning_rate": 4.333786675196539e-06,
+      "loss": 0.2494,
+      "step": 2087
+    },
+    {
+      "epoch": 2.7473684210526317,
+      "grad_norm": 0.08786933648344666,
+      "learning_rate": 4.2893163324085885e-06,
+      "loss": 0.2391,
+      "step": 2088
+    },
+    {
+      "epoch": 2.748684210526316,
+      "grad_norm": 0.0864833290624945,
+      "learning_rate": 4.245070328923584e-06,
+      "loss": 0.2368,
+      "step": 2089
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.08633401270649953,
+      "learning_rate": 4.20104876845111e-06,
+      "loss": 0.2371,
+      "step": 2090
+    },
+    {
+      "epoch": 2.751315789473684,
+      "grad_norm": 0.0872957150222657,
+      "learning_rate": 4.1572517541747294e-06,
+      "loss": 0.2411,
+      "step": 2091
+    },
+    {
+      "epoch": 2.7526315789473683,
+      "grad_norm": 0.08503214350266516,
+      "learning_rate": 4.1136793887516345e-06,
+      "loss": 0.2208,
+      "step": 2092
+    },
+    {
+      "epoch": 2.7539473684210525,
+      "grad_norm": 0.08709188346280049,
+      "learning_rate": 4.070331774312486e-06,
+      "loss": 0.2518,
+      "step": 2093
+    },
+    {
+      "epoch": 2.7552631578947366,
+      "grad_norm": 0.08684645991108973,
+      "learning_rate": 4.027209012461108e-06,
+      "loss": 0.2442,
+      "step": 2094
+    },
+    {
+      "epoch": 2.7565789473684212,
+      "grad_norm": 0.0857591344523911,
+      "learning_rate": 3.9843112042743045e-06,
+      "loss": 0.2321,
+      "step": 2095
+    },
+    {
+      "epoch": 2.7578947368421054,
+      "grad_norm": 0.08806771319356137,
+      "learning_rate": 3.941638450301644e-06,
+      "loss": 0.2497,
+      "step": 2096
+    },
+    {
+      "epoch": 2.7592105263157896,
+      "grad_norm": 0.08562267017890308,
+      "learning_rate": 3.899190850565115e-06,
+      "loss": 0.2305,
+      "step": 2097
+    },
+    {
+      "epoch": 2.7605263157894737,
+      "grad_norm": 0.0875801485743485,
+      "learning_rate": 3.856968504558989e-06,
+      "loss": 0.2461,
+      "step": 2098
+    },
+    {
+      "epoch": 2.761842105263158,
+      "grad_norm": 0.086786356254571,
+      "learning_rate": 3.814971511249576e-06,
+      "loss": 0.2399,
+      "step": 2099
+    },
+    {
+      "epoch": 2.763157894736842,
+      "grad_norm": 0.08869720806043141,
+      "learning_rate": 3.7731999690749585e-06,
+      "loss": 0.2418,
+      "step": 2100
+    },
+    {
+      "epoch": 2.764473684210526,
+      "grad_norm": 0.08740038989320052,
+      "learning_rate": 3.731653975944782e-06,
+      "loss": 0.2313,
+      "step": 2101
+    },
+    {
+      "epoch": 2.765789473684211,
+      "grad_norm": 0.08421195141482454,
+      "learning_rate": 3.690333629239995e-06,
+      "loss": 0.2452,
+      "step": 2102
+    },
+    {
+      "epoch": 2.7671052631578945,
+      "grad_norm": 0.08786793240306752,
+      "learning_rate": 3.6492390258126673e-06,
+      "loss": 0.2447,
+      "step": 2103
+    },
+    {
+      "epoch": 2.768421052631579,
+      "grad_norm": 0.08677843897373576,
+      "learning_rate": 3.6083702619857605e-06,
+      "loss": 0.2302,
+      "step": 2104
+    },
+    {
+      "epoch": 2.7697368421052633,
+      "grad_norm": 0.08762759404913377,
+      "learning_rate": 3.567727433552859e-06,
+      "loss": 0.2398,
+      "step": 2105
+    },
+    {
+      "epoch": 2.7710526315789474,
+      "grad_norm": 0.08696670352877477,
+      "learning_rate": 3.5273106357779585e-06,
+      "loss": 0.2303,
+      "step": 2106
+    },
+    {
+      "epoch": 2.7723684210526316,
+      "grad_norm": 0.08376635095287631,
+      "learning_rate": 3.4871199633953024e-06,
+      "loss": 0.2312,
+      "step": 2107
+    },
+    {
+      "epoch": 2.7736842105263158,
+      "grad_norm": 0.08618277832613587,
+      "learning_rate": 3.447155510609057e-06,
+      "loss": 0.2297,
+      "step": 2108
+    },
+    {
+      "epoch": 2.775,
+      "grad_norm": 0.0846951369483359,
+      "learning_rate": 3.40741737109318e-06,
+      "loss": 0.2277,
+      "step": 2109
+    },
+    {
+      "epoch": 2.776315789473684,
+      "grad_norm": 0.08571322444230144,
+      "learning_rate": 3.367905637991142e-06,
+      "loss": 0.235,
+      "step": 2110
+    },
+    {
+      "epoch": 2.7776315789473687,
+      "grad_norm": 0.08587956669273608,
+      "learning_rate": 3.328620403915761e-06,
+      "loss": 0.2309,
+      "step": 2111
+    },
+    {
+      "epoch": 2.7789473684210524,
+      "grad_norm": 0.08783109845186413,
+      "learning_rate": 3.2895617609489336e-06,
+      "loss": 0.2297,
+      "step": 2112
+    },
+    {
+      "epoch": 2.780263157894737,
+      "grad_norm": 0.08554797023694241,
+      "learning_rate": 3.2507298006414497e-06,
+      "loss": 0.235,
+      "step": 2113
+    },
+    {
+      "epoch": 2.781578947368421,
+      "grad_norm": 0.08680930677926671,
+      "learning_rate": 3.212124614012768e-06,
+      "loss": 0.2211,
+      "step": 2114
+    },
+    {
+      "epoch": 2.7828947368421053,
+      "grad_norm": 0.08766866674749912,
+      "learning_rate": 3.1737462915508277e-06,
+      "loss": 0.2337,
+      "step": 2115
+    },
+    {
+      "epoch": 2.7842105263157895,
+      "grad_norm": 0.08826029132090077,
+      "learning_rate": 3.135594923211771e-06,
+      "loss": 0.2422,
+      "step": 2116
+    },
+    {
+      "epoch": 2.7855263157894736,
+      "grad_norm": 0.08636371899570172,
+      "learning_rate": 3.0976705984198106e-06,
+      "loss": 0.2338,
+      "step": 2117
+    },
+    {
+      "epoch": 2.786842105263158,
+      "grad_norm": 0.09050217986657871,
+      "learning_rate": 3.059973406066963e-06,
+      "loss": 0.228,
+      "step": 2118
+    },
+    {
+      "epoch": 2.788157894736842,
+      "grad_norm": 0.08503847802029245,
+      "learning_rate": 3.02250343451288e-06,
+      "loss": 0.2271,
+      "step": 2119
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.08703828648807459,
+      "learning_rate": 2.9852607715846193e-06,
+      "loss": 0.2366,
+      "step": 2120
+    },
+    {
+      "epoch": 2.7907894736842103,
+      "grad_norm": 0.08829271415423992,
+      "learning_rate": 2.948245504576419e-06,
+      "loss": 0.2426,
+      "step": 2121
+    },
+    {
+      "epoch": 2.792105263157895,
+      "grad_norm": 0.08927594751571939,
+      "learning_rate": 2.9114577202495553e-06,
+      "loss": 0.2404,
+      "step": 2122
+    },
+    {
+      "epoch": 2.793421052631579,
+      "grad_norm": 0.08694095520332024,
+      "learning_rate": 2.874897504832075e-06,
+      "loss": 0.2378,
+      "step": 2123
+    },
+    {
+      "epoch": 2.794736842105263,
+      "grad_norm": 0.08685404680967446,
+      "learning_rate": 2.838564944018618e-06,
+      "loss": 0.2321,
+      "step": 2124
+    },
+    {
+      "epoch": 2.7960526315789473,
+      "grad_norm": 0.08768289772849402,
+      "learning_rate": 2.802460122970241e-06,
+      "loss": 0.241,
+      "step": 2125
+    },
+    {
+      "epoch": 2.7973684210526315,
+      "grad_norm": 0.08607902569363003,
+      "learning_rate": 2.7665831263141593e-06,
+      "loss": 0.233,
+      "step": 2126
+    },
+    {
+      "epoch": 2.7986842105263157,
+      "grad_norm": 0.08510692543767813,
+      "learning_rate": 2.730934038143607e-06,
+      "loss": 0.2356,
+      "step": 2127
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.08529239376866304,
+      "learning_rate": 2.6955129420176196e-06,
+      "loss": 0.2368,
+      "step": 2128
+    },
+    {
+      "epoch": 2.8013157894736844,
+      "grad_norm": 0.08535080650525781,
+      "learning_rate": 2.6603199209608187e-06,
+      "loss": 0.225,
+      "step": 2129
+    },
+    {
+      "epoch": 2.8026315789473686,
+      "grad_norm": 0.08544159810069368,
+      "learning_rate": 2.6253550574632303e-06,
+      "loss": 0.2278,
+      "step": 2130
+    },
+    {
+      "epoch": 2.8039473684210527,
+      "grad_norm": 0.0864151284971542,
+      "learning_rate": 2.5906184334801297e-06,
+      "loss": 0.24,
+      "step": 2131
+    },
+    {
+      "epoch": 2.805263157894737,
+      "grad_norm": 0.08574975605717337,
+      "learning_rate": 2.556110130431788e-06,
+      "loss": 0.2266,
+      "step": 2132
+    },
+    {
+      "epoch": 2.806578947368421,
+      "grad_norm": 0.0857571071786596,
+      "learning_rate": 2.5218302292032816e-06,
+      "loss": 0.234,
+      "step": 2133
+    },
+    {
+      "epoch": 2.807894736842105,
+      "grad_norm": 0.08886112337240165,
+      "learning_rate": 2.487778810144381e-06,
+      "loss": 0.2369,
+      "step": 2134
+    },
+    {
+      "epoch": 2.8092105263157894,
+      "grad_norm": 0.08877261222424743,
+      "learning_rate": 2.4539559530692758e-06,
+      "loss": 0.2304,
+      "step": 2135
+    },
+    {
+      "epoch": 2.8105263157894735,
+      "grad_norm": 0.08399044671282317,
+      "learning_rate": 2.420361737256438e-06,
+      "loss": 0.2282,
+      "step": 2136
+    },
+    {
+      "epoch": 2.8118421052631577,
+      "grad_norm": 0.08240687984849308,
+      "learning_rate": 2.3869962414484137e-06,
+      "loss": 0.2231,
+      "step": 2137
+    },
+    {
+      "epoch": 2.8131578947368423,
+      "grad_norm": 0.08767361332290152,
+      "learning_rate": 2.353859543851644e-06,
+      "loss": 0.2296,
+      "step": 2138
+    },
+    {
+      "epoch": 2.8144736842105265,
+      "grad_norm": 0.08790627521525603,
+      "learning_rate": 2.3209517221362777e-06,
+      "loss": 0.2293,
+      "step": 2139
+    },
+    {
+      "epoch": 2.8157894736842106,
+      "grad_norm": 0.08622647934206558,
+      "learning_rate": 2.288272853436013e-06,
+      "loss": 0.2325,
+      "step": 2140
+    },
+    {
+      "epoch": 2.817105263157895,
+      "grad_norm": 0.08713384140492961,
+      "learning_rate": 2.2558230143478797e-06,
+      "loss": 0.2276,
+      "step": 2141
+    },
+    {
+      "epoch": 2.818421052631579,
+      "grad_norm": 0.08415303494815114,
+      "learning_rate": 2.22360228093208e-06,
+      "loss": 0.221,
+      "step": 2142
+    },
+    {
+      "epoch": 2.819736842105263,
+      "grad_norm": 0.08558597543600141,
+      "learning_rate": 2.1916107287118015e-06,
+      "loss": 0.2293,
+      "step": 2143
+    },
+    {
+      "epoch": 2.8210526315789473,
+      "grad_norm": 0.08465218463761441,
+      "learning_rate": 2.1598484326730837e-06,
+      "loss": 0.2314,
+      "step": 2144
+    },
+    {
+      "epoch": 2.8223684210526314,
+      "grad_norm": 0.08774203550326841,
+      "learning_rate": 2.128315467264552e-06,
+      "loss": 0.2359,
+      "step": 2145
+    },
+    {
+      "epoch": 2.8236842105263156,
+      "grad_norm": 0.08602355712846993,
+      "learning_rate": 2.097011906397339e-06,
+      "loss": 0.2414,
+      "step": 2146
+    },
+    {
+      "epoch": 2.825,
+      "grad_norm": 0.09054024554229902,
+      "learning_rate": 2.0659378234448525e-06,
+      "loss": 0.2335,
+      "step": 2147
+    },
+    {
+      "epoch": 2.8263157894736843,
+      "grad_norm": 0.08533131567031418,
+      "learning_rate": 2.035093291242607e-06,
+      "loss": 0.2343,
+      "step": 2148
+    },
+    {
+      "epoch": 2.8276315789473685,
+      "grad_norm": 0.0884644294438792,
+      "learning_rate": 2.004478382088093e-06,
+      "loss": 0.2345,
+      "step": 2149
+    },
+    {
+      "epoch": 2.8289473684210527,
+      "grad_norm": 0.08947476508746073,
+      "learning_rate": 1.974093167740565e-06,
+      "loss": 0.2335,
+      "step": 2150
+    },
+    {
+      "epoch": 2.830263157894737,
+      "grad_norm": 0.086504450962513,
+      "learning_rate": 1.943937719420863e-06,
+      "loss": 0.2414,
+      "step": 2151
+    },
+    {
+      "epoch": 2.831578947368421,
+      "grad_norm": 0.08547620700378816,
+      "learning_rate": 1.914012107811336e-06,
+      "loss": 0.2304,
+      "step": 2152
+    },
+    {
+      "epoch": 2.832894736842105,
+      "grad_norm": 0.08406046395791139,
+      "learning_rate": 1.8843164030555527e-06,
+      "loss": 0.2224,
+      "step": 2153
+    },
+    {
+      "epoch": 2.8342105263157897,
+      "grad_norm": 0.0858958830300312,
+      "learning_rate": 1.8548506747582129e-06,
+      "loss": 0.2299,
+      "step": 2154
+    },
+    {
+      "epoch": 2.8355263157894735,
+      "grad_norm": 0.08713981466515258,
+      "learning_rate": 1.8256149919849807e-06,
+      "loss": 0.24,
+      "step": 2155
+    },
+    {
+      "epoch": 2.836842105263158,
+      "grad_norm": 0.08722563464111789,
+      "learning_rate": 1.7966094232622855e-06,
+      "loss": 0.2369,
+      "step": 2156
+    },
+    {
+      "epoch": 2.838157894736842,
+      "grad_norm": 0.08782546599587424,
+      "learning_rate": 1.7678340365772206e-06,
+      "loss": 0.2339,
+      "step": 2157
+    },
+    {
+      "epoch": 2.8394736842105264,
+      "grad_norm": 0.08820898856927462,
+      "learning_rate": 1.7392888993773005e-06,
+      "loss": 0.2294,
+      "step": 2158
+    },
+    {
+      "epoch": 2.8407894736842105,
+      "grad_norm": 0.08559547137262101,
+      "learning_rate": 1.7109740785703933e-06,
+      "loss": 0.233,
+      "step": 2159
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.08651933564372974,
+      "learning_rate": 1.6828896405244988e-06,
+      "loss": 0.2326,
+      "step": 2160
+    },
+    {
+      "epoch": 2.843421052631579,
+      "grad_norm": 0.08423235088748676,
+      "learning_rate": 1.6550356510676268e-06,
+      "loss": 0.2314,
+      "step": 2161
+    },
+    {
+      "epoch": 2.844736842105263,
+      "grad_norm": 0.08646992121393281,
+      "learning_rate": 1.6274121754876082e-06,
+      "loss": 0.2358,
+      "step": 2162
+    },
+    {
+      "epoch": 2.8460526315789476,
+      "grad_norm": 0.08897157982270418,
+      "learning_rate": 1.6000192785320057e-06,
+      "loss": 0.2421,
+      "step": 2163
+    },
+    {
+      "epoch": 2.8473684210526313,
+      "grad_norm": 0.08753330536214494,
+      "learning_rate": 1.572857024407881e-06,
+      "loss": 0.2367,
+      "step": 2164
+    },
+    {
+      "epoch": 2.848684210526316,
+      "grad_norm": 0.08917423233859062,
+      "learning_rate": 1.5459254767817066e-06,
+      "loss": 0.2337,
+      "step": 2165
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.08718844159110516,
+      "learning_rate": 1.5192246987791981e-06,
+      "loss": 0.2316,
+      "step": 2166
+    },
+    {
+      "epoch": 2.8513157894736842,
+      "grad_norm": 0.08666099472117271,
+      "learning_rate": 1.4927547529851371e-06,
+      "loss": 0.2145,
+      "step": 2167
+    },
+    {
+      "epoch": 2.8526315789473684,
+      "grad_norm": 0.08442460380559827,
+      "learning_rate": 1.466515701443294e-06,
+      "loss": 0.2332,
+      "step": 2168
+    },
+    {
+      "epoch": 2.8539473684210526,
+      "grad_norm": 0.08449859016985016,
+      "learning_rate": 1.4405076056561828e-06,
+      "loss": 0.2266,
+      "step": 2169
+    },
+    {
+      "epoch": 2.8552631578947367,
+      "grad_norm": 0.08545748600057546,
+      "learning_rate": 1.4147305265850175e-06,
+      "loss": 0.2298,
+      "step": 2170
+    },
+    {
+      "epoch": 2.856578947368421,
+      "grad_norm": 0.08929470679144376,
+      "learning_rate": 1.3891845246495228e-06,
+      "loss": 0.2343,
+      "step": 2171
+    },
+    {
+      "epoch": 2.8578947368421055,
+      "grad_norm": 0.08668972660138131,
+      "learning_rate": 1.3638696597277679e-06,
+      "loss": 0.2423,
+      "step": 2172
+    },
+    {
+      "epoch": 2.859210526315789,
+      "grad_norm": 0.08262056937095345,
+      "learning_rate": 1.3387859911560664e-06,
+      "loss": 0.2285,
+      "step": 2173
+    },
+    {
+      "epoch": 2.860526315789474,
+      "grad_norm": 0.08830815458424673,
+      "learning_rate": 1.3139335777288208e-06,
+      "loss": 0.2334,
+      "step": 2174
+    },
+    {
+      "epoch": 2.861842105263158,
+      "grad_norm": 0.08675825238016463,
+      "learning_rate": 1.28931247769839e-06,
+      "loss": 0.2382,
+      "step": 2175
+    },
+    {
+      "epoch": 2.863157894736842,
+      "grad_norm": 0.08893171231303189,
+      "learning_rate": 1.2649227487749548e-06,
+      "loss": 0.2336,
+      "step": 2176
+    },
+    {
+      "epoch": 2.8644736842105263,
+      "grad_norm": 0.08924178196183445,
+      "learning_rate": 1.2407644481263858e-06,
+      "loss": 0.2408,
+      "step": 2177
+    },
+    {
+      "epoch": 2.8657894736842104,
+      "grad_norm": 0.08715372943476031,
+      "learning_rate": 1.216837632378065e-06,
+      "loss": 0.2315,
+      "step": 2178
+    },
+    {
+      "epoch": 2.8671052631578946,
+      "grad_norm": 0.08660959792084139,
+      "learning_rate": 1.1931423576128197e-06,
+      "loss": 0.2416,
+      "step": 2179
+    },
+    {
+      "epoch": 2.8684210526315788,
+      "grad_norm": 0.08853921306085291,
+      "learning_rate": 1.1696786793707781e-06,
+      "loss": 0.2335,
+      "step": 2180
+    },
+    {
+      "epoch": 2.8697368421052634,
+      "grad_norm": 0.08640447512392578,
+      "learning_rate": 1.146446652649169e-06,
+      "loss": 0.2358,
+      "step": 2181
+    },
+    {
+      "epoch": 2.8710526315789475,
+      "grad_norm": 0.08705591508512092,
+      "learning_rate": 1.1234463319022893e-06,
+      "loss": 0.2315,
+      "step": 2182
+    },
+    {
+      "epoch": 2.8723684210526317,
+      "grad_norm": 0.08602413068650967,
+      "learning_rate": 1.100677771041314e-06,
+      "loss": 0.2235,
+      "step": 2183
+    },
+    {
+      "epoch": 2.873684210526316,
+      "grad_norm": 0.08684507106024413,
+      "learning_rate": 1.0781410234342094e-06,
+      "loss": 0.24,
+      "step": 2184
+    },
+    {
+      "epoch": 2.875,
+      "grad_norm": 0.09031170729723671,
+      "learning_rate": 1.055836141905553e-06,
+      "loss": 0.2338,
+      "step": 2185
+    },
+    {
+      "epoch": 2.876315789473684,
+      "grad_norm": 0.08646029200839189,
+      "learning_rate": 1.0337631787364687e-06,
+      "loss": 0.2222,
+      "step": 2186
+    },
+    {
+      "epoch": 2.8776315789473683,
+      "grad_norm": 0.08740568430137426,
+      "learning_rate": 1.0119221856644712e-06,
+      "loss": 0.2352,
+      "step": 2187
+    },
+    {
+      "epoch": 2.8789473684210525,
+      "grad_norm": 0.08586656965987803,
+      "learning_rate": 9.90313213883376e-07,
+      "loss": 0.2263,
+      "step": 2188
+    },
+    {
+      "epoch": 2.8802631578947366,
+      "grad_norm": 0.08991958053706495,
+      "learning_rate": 9.689363140431118e-07,
+      "loss": 0.2266,
+      "step": 2189
+    },
+    {
+      "epoch": 2.8815789473684212,
+      "grad_norm": 0.08767297790926798,
+      "learning_rate": 9.477915362496758e-07,
+      "loss": 0.2441,
+      "step": 2190
+    },
+    {
+      "epoch": 2.8828947368421054,
+      "grad_norm": 0.0898797640991881,
+      "learning_rate": 9.268789300649894e-07,
+      "loss": 0.2401,
+      "step": 2191
+    },
+    {
+      "epoch": 2.8842105263157896,
+      "grad_norm": 0.08612849691104628,
+      "learning_rate": 9.061985445067756e-07,
+      "loss": 0.2281,
+      "step": 2192
+    },
+    {
+      "epoch": 2.8855263157894737,
+      "grad_norm": 0.08949331264552439,
+      "learning_rate": 8.857504280484375e-07,
+      "loss": 0.2547,
+      "step": 2193
+    },
+    {
+      "epoch": 2.886842105263158,
+      "grad_norm": 0.0851227718307587,
+      "learning_rate": 8.65534628618958e-07,
+      "loss": 0.2297,
+      "step": 2194
+    },
+    {
+      "epoch": 2.888157894736842,
+      "grad_norm": 0.08827554977188867,
+      "learning_rate": 8.455511936028004e-07,
+      "loss": 0.2411,
+      "step": 2195
+    },
+    {
+      "epoch": 2.889473684210526,
+      "grad_norm": 0.08846981462555736,
+      "learning_rate": 8.258001698397744e-07,
+      "loss": 0.2389,
+      "step": 2196
+    },
+    {
+      "epoch": 2.890789473684211,
+      "grad_norm": 0.08664150903305488,
+      "learning_rate": 8.062816036249143e-07,
+      "loss": 0.2345,
+      "step": 2197
+    },
+    {
+      "epoch": 2.8921052631578945,
+      "grad_norm": 0.0854396346836948,
+      "learning_rate": 7.86995540708424e-07,
+      "loss": 0.2318,
+      "step": 2198
+    },
+    {
+      "epoch": 2.893421052631579,
+      "grad_norm": 0.08695987453269646,
+      "learning_rate": 7.679420262954984e-07,
+      "loss": 0.2424,
+      "step": 2199
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.08752689227211352,
+      "learning_rate": 7.491211050462798e-07,
+      "loss": 0.2304,
+      "step": 2200
+    },
+    {
+      "epoch": 2.8960526315789474,
+      "grad_norm": 0.08666285275563744,
+      "learning_rate": 7.305328210757356e-07,
+      "loss": 0.2299,
+      "step": 2201
+    },
+    {
+      "epoch": 2.8973684210526316,
+      "grad_norm": 0.08663314353560102,
+      "learning_rate": 7.121772179535135e-07,
+      "loss": 0.2424,
+      "step": 2202
+    },
+    {
+      "epoch": 2.8986842105263158,
+      "grad_norm": 0.08458000982804612,
+      "learning_rate": 6.94054338703909e-07,
+      "loss": 0.2316,
+      "step": 2203
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.08821934440133324,
+      "learning_rate": 6.761642258056978e-07,
+      "loss": 0.2276,
+      "step": 2204
+    },
+    {
+      "epoch": 2.901315789473684,
+      "grad_norm": 0.08815511451469815,
+      "learning_rate": 6.585069211921035e-07,
+      "loss": 0.2379,
+      "step": 2205
+    },
+    {
+      "epoch": 2.9026315789473687,
+      "grad_norm": 0.08529196165876289,
+      "learning_rate": 6.410824662506198e-07,
+      "loss": 0.2296,
+      "step": 2206
+    },
+    {
+      "epoch": 2.9039473684210524,
+      "grad_norm": 0.08451900774747255,
+      "learning_rate": 6.238909018229766e-07,
+      "loss": 0.2269,
+      "step": 2207
+    },
+    {
+      "epoch": 2.905263157894737,
+      "grad_norm": 0.08696348620201358,
+      "learning_rate": 6.069322682050516e-07,
+      "loss": 0.2324,
+      "step": 2208
+    },
+    {
+      "epoch": 2.906578947368421,
+      "grad_norm": 0.08828910381160303,
+      "learning_rate": 5.902066051467037e-07,
+      "loss": 0.2357,
+      "step": 2209
+    },
+    {
+      "epoch": 2.9078947368421053,
+      "grad_norm": 0.08483736507293088,
+      "learning_rate": 5.737139518517509e-07,
+      "loss": 0.2296,
+      "step": 2210
+    },
+    {
+      "epoch": 2.9092105263157895,
+      "grad_norm": 0.08743099417381292,
+      "learning_rate": 5.57454346977837e-07,
+      "loss": 0.24,
+      "step": 2211
+    },
+    {
+      "epoch": 2.9105263157894736,
+      "grad_norm": 0.08916802794242618,
+      "learning_rate": 5.414278286363761e-07,
+      "loss": 0.2304,
+      "step": 2212
+    },
+    {
+      "epoch": 2.911842105263158,
+      "grad_norm": 0.08828220707425087,
+      "learning_rate": 5.256344343924302e-07,
+      "loss": 0.2419,
+      "step": 2213
+    },
+    {
+      "epoch": 2.913157894736842,
+      "grad_norm": 0.08659025242576524,
+      "learning_rate": 5.10074201264632e-07,
+      "loss": 0.2166,
+      "step": 2214
+    },
+    {
+      "epoch": 2.9144736842105265,
+      "grad_norm": 0.08525185958778574,
+      "learning_rate": 4.947471657251068e-07,
+      "loss": 0.2368,
+      "step": 2215
+    },
+    {
+      "epoch": 2.9157894736842103,
+      "grad_norm": 0.08692863494788813,
+      "learning_rate": 4.796533636993727e-07,
+      "loss": 0.23,
+      "step": 2216
+    },
+    {
+      "epoch": 2.917105263157895,
+      "grad_norm": 0.08819398487450925,
+      "learning_rate": 4.647928305662852e-07,
+      "loss": 0.2321,
+      "step": 2217
+    },
+    {
+      "epoch": 2.918421052631579,
+      "grad_norm": 0.08936510229812145,
+      "learning_rate": 4.501656011579036e-07,
+      "loss": 0.2468,
+      "step": 2218
+    },
+    {
+      "epoch": 2.919736842105263,
+      "grad_norm": 0.08708818005334107,
+      "learning_rate": 4.3577170975945826e-07,
+      "loss": 0.2322,
+      "step": 2219
+    },
+    {
+      "epoch": 2.9210526315789473,
+      "grad_norm": 0.08388217210536535,
+      "learning_rate": 4.216111901092501e-07,
+      "loss": 0.2267,
+      "step": 2220
+    },
+    {
+      "epoch": 2.9223684210526315,
+      "grad_norm": 0.08876373477705209,
+      "learning_rate": 4.0768407539857333e-07,
+      "loss": 0.2337,
+      "step": 2221
+    },
+    {
+      "epoch": 2.9236842105263157,
+      "grad_norm": 0.08918972976359953,
+      "learning_rate": 3.9399039827162643e-07,
+      "loss": 0.2458,
+      "step": 2222
+    },
+    {
+      "epoch": 2.925,
+      "grad_norm": 0.08612029906467318,
+      "learning_rate": 3.805301908254455e-07,
+      "loss": 0.2243,
+      "step": 2223
+    },
+    {
+      "epoch": 2.9263157894736844,
+      "grad_norm": 0.08294679916711059,
+      "learning_rate": 3.6730348460985996e-07,
+      "loss": 0.2184,
+      "step": 2224
+    },
+    {
+      "epoch": 2.9276315789473686,
+      "grad_norm": 0.08752162048000231,
+      "learning_rate": 3.543103106273371e-07,
+      "loss": 0.2457,
+      "step": 2225
+    },
+    {
+      "epoch": 2.9289473684210527,
+      "grad_norm": 0.0859025435788662,
+      "learning_rate": 3.415506993330153e-07,
+      "loss": 0.2346,
+      "step": 2226
+    },
+    {
+      "epoch": 2.930263157894737,
+      "grad_norm": 0.08793876747475723,
+      "learning_rate": 3.2902468063453763e-07,
+      "loss": 0.2391,
+      "step": 2227
+    },
+    {
+      "epoch": 2.931578947368421,
+      "grad_norm": 0.08667819600458836,
+      "learning_rate": 3.1673228389204055e-07,
+      "loss": 0.2288,
+      "step": 2228
+    },
+    {
+      "epoch": 2.932894736842105,
+      "grad_norm": 0.08786519232774889,
+      "learning_rate": 3.046735379180543e-07,
+      "loss": 0.2341,
+      "step": 2229
+    },
+    {
+      "epoch": 2.9342105263157894,
+      "grad_norm": 0.0901967616524326,
+      "learning_rate": 2.9284847097746923e-07,
+      "loss": 0.2291,
+      "step": 2230
+    },
+    {
+      "epoch": 2.9355263157894735,
+      "grad_norm": 0.08704968172092865,
+      "learning_rate": 2.81257110787414e-07,
+      "loss": 0.234,
+      "step": 2231
+    },
+    {
+      "epoch": 2.9368421052631577,
+      "grad_norm": 0.08690111626726336,
+      "learning_rate": 2.6989948451726643e-07,
+      "loss": 0.2284,
+      "step": 2232
+    },
+    {
+      "epoch": 2.9381578947368423,
+      "grad_norm": 0.08848680506181238,
+      "learning_rate": 2.587756187885204e-07,
+      "loss": 0.2313,
+      "step": 2233
+    },
+    {
+      "epoch": 2.9394736842105265,
+      "grad_norm": 0.08596311834903651,
+      "learning_rate": 2.4788553967474147e-07,
+      "loss": 0.2361,
+      "step": 2234
+    },
+    {
+      "epoch": 2.9407894736842106,
+      "grad_norm": 0.08490908394075776,
+      "learning_rate": 2.372292727015557e-07,
+      "loss": 0.2353,
+      "step": 2235
+    },
+    {
+      "epoch": 2.942105263157895,
+      "grad_norm": 0.08292706658982998,
+      "learning_rate": 2.2680684284650533e-07,
+      "loss": 0.2309,
+      "step": 2236
+    },
+    {
+      "epoch": 2.943421052631579,
+      "grad_norm": 0.08480952928040106,
+      "learning_rate": 2.1661827453905992e-07,
+      "loss": 0.2213,
+      "step": 2237
+    },
+    {
+      "epoch": 2.944736842105263,
+      "grad_norm": 0.08635227023585973,
+      "learning_rate": 2.066635916605386e-07,
+      "loss": 0.2353,
+      "step": 2238
+    },
+    {
+      "epoch": 2.9460526315789473,
+      "grad_norm": 0.08781565402914938,
+      "learning_rate": 1.9694281754401024e-07,
+      "loss": 0.2361,
+      "step": 2239
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.08776480608964969,
+      "learning_rate": 1.8745597497433765e-07,
+      "loss": 0.2415,
+      "step": 2240
+    },
+    {
+      "epoch": 2.9486842105263156,
+      "grad_norm": 0.08905978886837364,
+      "learning_rate": 1.782030861880113e-07,
+      "loss": 0.2308,
+      "step": 2241
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.08890792718531305,
+      "learning_rate": 1.6918417287318245e-07,
+      "loss": 0.2415,
+      "step": 2242
+    },
+    {
+      "epoch": 2.9513157894736843,
+      "grad_norm": 0.08905016422570772,
+      "learning_rate": 1.603992561695522e-07,
+      "loss": 0.2378,
+      "step": 2243
+    },
+    {
+      "epoch": 2.9526315789473685,
+      "grad_norm": 0.08893795963848528,
+      "learning_rate": 1.518483566683826e-07,
+      "loss": 0.2435,
+      "step": 2244
+    },
+    {
+      "epoch": 2.9539473684210527,
+      "grad_norm": 0.08778434845315054,
+      "learning_rate": 1.4353149441237445e-07,
+      "loss": 0.2378,
+      "step": 2245
+    },
+    {
+      "epoch": 2.955263157894737,
+      "grad_norm": 0.08482607245940768,
+      "learning_rate": 1.3544868889571182e-07,
+      "loss": 0.2206,
+      "step": 2246
+    },
+    {
+      "epoch": 2.956578947368421,
+      "grad_norm": 0.08524277713713192,
+      "learning_rate": 1.2759995906392874e-07,
+      "loss": 0.2235,
+      "step": 2247
+    },
+    {
+      "epoch": 2.957894736842105,
+      "grad_norm": 0.08760984571309041,
+      "learning_rate": 1.199853233138981e-07,
+      "loss": 0.2301,
+      "step": 2248
+    },
+    {
+      "epoch": 2.9592105263157897,
+      "grad_norm": 0.08864338113540858,
+      "learning_rate": 1.1260479949382064e-07,
+      "loss": 0.241,
+      "step": 2249
+    },
+    {
+      "epoch": 2.9605263157894735,
+      "grad_norm": 0.0860770469372616,
+      "learning_rate": 1.0545840490313596e-07,
+      "loss": 0.238,
+      "step": 2250
+    },
+    {
+      "epoch": 2.961842105263158,
+      "grad_norm": 0.08309356073768617,
+      "learning_rate": 9.854615629250053e-08,
+      "loss": 0.2216,
+      "step": 2251
+    },
+    {
+      "epoch": 2.963157894736842,
+      "grad_norm": 0.08693365254726673,
+      "learning_rate": 9.186806986376529e-08,
+      "loss": 0.229,
+      "step": 2252
+    },
+    {
+      "epoch": 2.9644736842105264,
+      "grad_norm": 0.08720773613369974,
+      "learning_rate": 8.542416126989805e-08,
+      "loss": 0.2323,
+      "step": 2253
+    },
+    {
+      "epoch": 2.9657894736842105,
+      "grad_norm": 0.089536393572043,
+      "learning_rate": 7.921444561498348e-08,
+      "loss": 0.2369,
+      "step": 2254
+    },
+    {
+      "epoch": 2.9671052631578947,
+      "grad_norm": 0.08652678030587639,
+      "learning_rate": 7.323893745416755e-08,
+      "loss": 0.2398,
+      "step": 2255
+    },
+    {
+      "epoch": 2.968421052631579,
+      "grad_norm": 0.08812917248570945,
+      "learning_rate": 6.749765079363534e-08,
+      "loss": 0.2262,
+      "step": 2256
+    },
+    {
+      "epoch": 2.969736842105263,
+      "grad_norm": 0.08805384258384784,
+      "learning_rate": 6.19905990905667e-08,
+      "loss": 0.2264,
+      "step": 2257
+    },
+    {
+      "epoch": 2.9710526315789476,
+      "grad_norm": 0.0845281113628984,
+      "learning_rate": 5.6717795253113935e-08,
+      "loss": 0.2296,
+      "step": 2258
+    },
+    {
+      "epoch": 2.9723684210526313,
+      "grad_norm": 0.08972586281120497,
+      "learning_rate": 5.167925164037968e-08,
+      "loss": 0.2444,
+      "step": 2259
+    },
+    {
+      "epoch": 2.973684210526316,
+      "grad_norm": 0.086226803286329,
+      "learning_rate": 4.687498006236135e-08,
+      "loss": 0.2377,
+      "step": 2260
+    },
+    {
+      "epoch": 2.975,
+      "grad_norm": 0.08827582008385458,
+      "learning_rate": 4.230499177994007e-08,
+      "loss": 0.2275,
+      "step": 2261
+    },
+    {
+      "epoch": 2.9763157894736842,
+      "grad_norm": 0.08805311064331064,
+      "learning_rate": 3.796929750485845e-08,
+      "loss": 0.2453,
+      "step": 2262
+    },
+    {
+      "epoch": 2.9776315789473684,
+      "grad_norm": 0.08707716445589411,
+      "learning_rate": 3.386790739968726e-08,
+      "loss": 0.2484,
+      "step": 2263
+    },
+    {
+      "epoch": 2.9789473684210526,
+      "grad_norm": 0.08937493080093048,
+      "learning_rate": 3.000083107780327e-08,
+      "loss": 0.2341,
+      "step": 2264
+    },
+    {
+      "epoch": 2.9802631578947367,
+      "grad_norm": 0.08847213274992714,
+      "learning_rate": 2.6368077603367015e-08,
+      "loss": 0.2345,
+      "step": 2265
+    },
+    {
+      "epoch": 2.981578947368421,
+      "grad_norm": 0.08752540842663672,
+      "learning_rate": 2.2969655491311693e-08,
+      "loss": 0.231,
+      "step": 2266
+    },
+    {
+      "epoch": 2.9828947368421055,
+      "grad_norm": 0.08820844196032962,
+      "learning_rate": 1.980557270729877e-08,
+      "loss": 0.2403,
+      "step": 2267
+    },
+    {
+      "epoch": 2.984210526315789,
+      "grad_norm": 0.08848035935218941,
+      "learning_rate": 1.687583666772907e-08,
+      "loss": 0.2475,
+      "step": 2268
+    },
+    {
+      "epoch": 2.985526315789474,
+      "grad_norm": 0.08684909478055165,
+      "learning_rate": 1.418045423968728e-08,
+      "loss": 0.2276,
+      "step": 2269
+    },
+    {
+      "epoch": 2.986842105263158,
+      "grad_norm": 0.08857801328002969,
+      "learning_rate": 1.1719431740997433e-08,
+      "loss": 0.2351,
+      "step": 2270
+    },
+    {
+      "epoch": 2.988157894736842,
+      "grad_norm": 0.08626632234298558,
+      "learning_rate": 9.49277494008971e-09,
+      "loss": 0.2371,
+      "step": 2271
+    },
+    {
+      "epoch": 2.9894736842105263,
+      "grad_norm": 0.08631453252565319,
+      "learning_rate": 7.500489056133652e-09,
+      "loss": 0.241,
+      "step": 2272
+    },
+    {
+      "epoch": 2.9907894736842104,
+      "grad_norm": 0.08837237273152578,
+      "learning_rate": 5.742578758882733e-09,
+      "loss": 0.2304,
+      "step": 2273
+    },
+    {
+      "epoch": 2.9921052631578946,
+      "grad_norm": 0.0862791762664004,
+      "learning_rate": 4.219048168763174e-09,
+      "loss": 0.2425,
+      "step": 2274
+    },
+    {
+      "epoch": 2.9934210526315788,
+      "grad_norm": 0.08930817551049701,
+      "learning_rate": 2.9299008568406396e-09,
+      "loss": 0.2402,
+      "step": 2275
+    },
+    {
+      "epoch": 2.9947368421052634,
+      "grad_norm": 0.08856125619943322,
+      "learning_rate": 1.8751398447758306e-09,
+      "loss": 0.2358,
+      "step": 2276
+    },
+    {
+      "epoch": 2.9960526315789475,
+      "grad_norm": 0.08640583110457638,
+      "learning_rate": 1.0547676048688892e-09,
+      "loss": 0.2259,
+      "step": 2277
+    },
+    {
+      "epoch": 2.9973684210526317,
+      "grad_norm": 0.08504843454486558,
+      "learning_rate": 4.687860599927873e-10,
+      "loss": 0.2402,
+      "step": 2278
+    },
+    {
+      "epoch": 2.998684210526316,
+      "grad_norm": 0.08642252232900641,
+      "learning_rate": 1.1719658367104202e-10,
+      "loss": 0.2366,
+      "step": 2279
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.08478813908452047,
+      "learning_rate": 0.0,
+      "loss": 0.2337,
+      "step": 2280
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.2625243365764618,
+      "eval_runtime": 136.232,
+      "eval_samples_per_second": 37.568,
+      "eval_steps_per_second": 1.174,
+      "step": 2280
+    },
+    {
+      "epoch": 3.0,
+      "step": 2280,
+      "total_flos": 6.730550114202419e+17,
+      "train_loss": 0.28347440996583095,
+      "train_runtime": 21673.3063,
+      "train_samples_per_second": 13.46,
+      "train_steps_per_second": 0.105
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2280,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.730550114202419e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}